{"arcee-ai/trinity-large-preview":{"id":"arcee-ai/trinity-large-preview","name":"Trinity Large Preview","description":"Lightly post-trained 398B MoE chat model for creative work, long-context prompts, and tool-using agents","family":"trinity","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-01-27","last_updated":"2026-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":262144},"license":"OpenMDW-1.1","links":[{"label":"Model card","url":"https://huggingface.co/arcee-ai/Trinity-Large-Preview","type":"model_card"},{"label":"Announcement","url":"https://www.arcee.ai/blog/trinity-large","type":"announcement"},{"label":"License","url":"https://huggingface.co/arcee-ai/Trinity-Large-Preview/blob/main/LICENSE","type":"license"}],"weights":[{"label":"Hugging Face","url":"https://huggingface.co/arcee-ai/Trinity-Large-Preview","format":"safetensors"}]},"arcee-ai/trinity-nano-preview":{"id":"arcee-ai/trinity-nano-preview","name":"Trinity Nano Preview","description":"Experimental chat-tuned 6B MoE model with 1B active parameters for low-resource chat and instruction following","family":"trinity","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-12-01","last_updated":"2026-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"license":"OpenMDW-1.1","links":[{"label":"Model card","url":"https://huggingface.co/arcee-ai/Trinity-Nano-Preview","type":"model_card"},{"label":"Announcement","url":"https://www.arcee.ai/blog/the-trinity-manifesto","type":"announcement"},{"label":"License","url":"https://huggingface.co/arcee-ai/Trinity-Nano-Preview/blob/main/LICENSE","type":"license"}],"weights":[{"label":"Hugging Face","url":"https://huggingface.co/arcee-ai/Trinity-Nano-Preview","format":"safetensors"}]},"arcee-ai/trinity-large-thinking":{"id":"arcee-ai/trinity-large-thinking","name":"Trinity Large Thinking","description":"Reasoning-optimized 398B MoE agent model with extended thinking for long-horizon and multi-turn tool use","family":"trinity","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":262144},"license":"OpenMDW-1.1","links":[{"label":"Model card","url":"https://huggingface.co/arcee-ai/Trinity-Large-Thinking","type":"model_card"},{"label":"Announcement","url":"https://www.arcee.ai/blog/trinity-large-thinking","type":"announcement"},{"label":"License","url":"https://huggingface.co/arcee-ai/Trinity-Large-Thinking/blob/main/LICENSE","type":"license"}],"weights":[{"label":"Hugging Face","url":"https://huggingface.co/arcee-ai/Trinity-Large-Thinking","format":"safetensors"}]},"arcee-ai/trinity-mini":{"id":"arcee-ai/trinity-mini","name":"Trinity Mini","description":"Reasoning-tuned 26B MoE model with 3B active parameters for agents, tools, and multi-step workloads","family":"trinity","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2025-12-01","last_updated":"2026-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"license":"OpenMDW-1.1","links":[{"label":"Model card","url":"https://huggingface.co/arcee-ai/Trinity-Mini","type":"model_card"},{"label":"Announcement","url":"https://www.arcee.ai/blog/the-trinity-manifesto","type":"announcement"},{"label":"License","url":"https://huggingface.co/arcee-ai/Trinity-Mini/blob/main/LICENSE","type":"license"}],"weights":[{"label":"Hugging Face","url":"https://huggingface.co/arcee-ai/Trinity-Mini","format":"safetensors"}]},"zhipuai/autoglm-phone-9b-multilingual":{"id":"zhipuai/autoglm-phone-9b-multilingual","name":"AutoGLM Phone 9B Multilingual","description":"Multilingual visual model for operating smartphone interfaces and mobile agents","family":"glm","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-12-10","last_updated":"2025-12-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":65536},"license":"MIT","weights":[{"label":"Hugging Face","url":"https://huggingface.co/zai-org/AutoGLM-Phone-9B-Multilingual"}]},"zhipuai/glm-5v-turbo":{"id":"zhipuai/glm-5v-turbo","name":"GLM-5V-Turbo","description":"Fast GLM vision model for screenshots, documents, and multimodal agent tasks","family":"glm","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-04-01","last_updated":"2026-04-01","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":131072}},"zhipuai/glm-4.7":{"id":"zhipuai/glm-4.7","name":"GLM-4.7","description":"Mature GLM model for dependable coding, reasoning, and structured agent tasks","family":"glm","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-22","last_updated":"2025-12-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/zai-org/GLM-4.7"}],"benchmarks":[{"name":"SWE-Bench Verified","score":73.8,"metric":"resolved","source":"https://huggingface.co/zai-org/GLM-4.7"},{"name":"Terminal Bench 2.0","score":33.4,"metric":"score","source":"https://huggingface.co/zai-org/GLM-4.7"}]},"zhipuai/glm-4.5v":{"id":"zhipuai/glm-4.5v","name":"GLM-4.5V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-08-11","last_updated":"2025-08-11","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":64000,"output":16384},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/zai-org/GLM-4.5V"}],"benchmarks":[{"name":"Artificial Analysis Coding Index","score":10.9,"metric":"index","source":"https://openrouter.ai/z-ai/glm-4.5v/benchmarks","date":"2026-04-29"},{"name":"SciCode","score":22.1,"metric":"percent correct","source":"https://openrouter.ai/z-ai/glm-4.5v/benchmarks","date":"2026-04-29"},{"name":"Terminal-Bench Hard","score":5.3,"metric":"success rate","source":"https://openrouter.ai/z-ai/glm-4.5v/benchmarks","date":"2026-04-29"}]},"zhipuai/glm-4.5-air":{"id":"zhipuai/glm-4.5-air","name":"GLM-4.5-Air","description":"Lighter GLM-4.5 variant for fast coding assistance and cheaper agents","family":"glm-air","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/zai-org/GLM-4.5-Air"}],"benchmarks":[{"name":"Artificial Analysis Coding Index","score":23.8,"metric":"index","source":"https://openrouter.ai/z-ai/glm-4.5-air/benchmarks","date":"2026-05-30"},{"name":"SciCode","score":30.6,"metric":"percent correct","source":"https://openrouter.ai/z-ai/glm-4.5-air/benchmarks","date":"2026-05-30"},{"name":"Terminal-Bench Hard","score":20.5,"metric":"success rate","source":"https://openrouter.ai/z-ai/glm-4.5-air/benchmarks","date":"2026-05-30"}]},"zhipuai/glm-4.5-flash":{"id":"zhipuai/glm-4.5-flash","name":"GLM-4.5-Flash","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":98304}},"zhipuai/glm-4.7-flashx":{"id":"zhipuai/glm-4.7-flashx","name":"GLM-4.7-FlashX","description":"Efficient GLM model for fast reasoning, coding, and agent workflows","family":"glm-flash","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/zai-org/GLM-4.7-Flash"}]},"zhipuai/glm-4.6v":{"id":"zhipuai/glm-4.6v","name":"GLM-4.6V","description":"GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/zai-org/GLM-4.6V"}]},"zhipuai/glm-5.2":{"id":"zhipuai/glm-5.2","name":"GLM-5.2","description":"Open flagship GLM for long-horizon coding agents and million-token context work","family":"glm","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-06-13","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/zai-org/GLM-5.2"}],"benchmarks":[{"name":"SWE-Bench Pro","score":62.1,"metric":"resolve rate","source":"https://z.ai/blog/glm-5.2","date":"2026-06-16"},{"name":"Terminal-Bench","score":82.7,"metric":"success rate","harness":"Claude Code","version":"2.1","source":"https://z.ai/blog/glm-5.2","date":"2026-06-16"},{"name":"FrontierSWE","score":74.4,"metric":"dominance","source":"https://z.ai/blog/glm-5.2","date":"2026-06-16"},{"name":"Humanity's Last Exam","score":40.5,"metric":"accuracy","dataset":"text-only subset","source":"https://z.ai/blog/glm-5.2","date":"2026-06-16"},{"name":"Humanity's Last Exam","score":54.7,"metric":"accuracy","variant":"with tools","dataset":"text-only subset","source":"https://z.ai/blog/glm-5.2","date":"2026-06-16"},{"name":"CritPt","score":20.9,"metric":"accuracy","source":"https://z.ai/blog/glm-5.2","date":"2026-06-16"},{"name":"AIME","score":99.2,"metric":"accuracy","version":"2026","source":"https://z.ai/blog/glm-5.2","date":"2026-06-16"},{"name":"HMMT","score":94.4,"metric":"accuracy","version":"November 2025","source":"https://z.ai/blog/glm-5.2","date":"2026-06-16"},{"name":"HMMT","score":92.5,"metric":"accuracy","version":"February 2026","source":"https://z.ai/blog/glm-5.2","date":"2026-06-16"},{"name":"IMOAnswerBench","score":91,"metric":"accuracy","source":"https://z.ai/blog/glm-5.2","date":"2026-06-16"},{"name":"GPQA Diamond","score":91.2,"metric":"accuracy","source":"https://z.ai/blog/glm-5.2","date":"2026-06-16"},{"name":"NL2Repo","score":48.9,"metric":"resolve rate","source":"https://z.ai/blog/glm-5.2","date":"2026-06-16"},{"name":"DeepSWE","score":46.2,"metric":"resolve rate","source":"https://z.ai/blog/glm-5.2","date":"2026-06-16"},{"name":"Program Bench","score":63.7,"metric":"score","source":"https://z.ai/blog/glm-5.2","date":"2026-06-16"},{"name":"Terminal-Bench","score":81,"metric":"success rate","harness":"Terminus 2","version":"2.1","source":"https://z.ai/blog/glm-5.2","date":"2026-06-16"},{"name":"PostTrainBench","score":34.3,"metric":"score","source":"https://z.ai/blog/glm-5.2","date":"2026-06-16"},{"name":"SWE Marathon","score":13,"metric":"resolve rate","source":"https://z.ai/blog/glm-5.2","date":"2026-06-16"},{"name":"MCP Atlas","score":76.8,"metric":"score","dataset":"public subset","source":"https://z.ai/blog/glm-5.2","date":"2026-06-16"},{"name":"Tool-Decathlon","score":48.2,"metric":"score","source":"https://z.ai/blog/glm-5.2","date":"2026-06-16"}]},"zhipuai/glm-5.3":{"id":"zhipuai/glm-5.3","name":"GLM-5.3","description":"Flagship GLM model for long-horizon coding, agents, and complex project delivery","family":"glm","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/zai-org/GLM-5.3"}],"benchmarks":[{"name":"Terminal-Bench","score":88.2,"metric":"pass@1","harness":"Claude Code 2.1.207","variant":"max effort","version":"2.1","source":"https://huggingface.co/zai-org/GLM-5.3"},{"name":"Terminal-Bench","score":28.3,"metric":"avg@3","harness":"Claude Code 2.1.207","variant":"max effort","version":"3.0","source":"https://huggingface.co/zai-org/GLM-5.3"},{"name":"DeepSWE","score":66.9,"metric":"resolved","harness":"mini-swe-agent","variant":"max effort; 6h timeout; 400K context","version":"1.1","source":"https://huggingface.co/zai-org/GLM-5.3"},{"name":"NL2Repo","score":58,"metric":"score","variant":"max effort","source":"https://huggingface.co/zai-org/GLM-5.3"},{"name":"ProgramBench","score":19,"metric":"almost solved","variant":"max effort","source":"https://huggingface.co/zai-org/GLM-5.3"},{"name":"FrontierSWE","score":78.1,"metric":"dominance score","variant":"max effort","source":"https://huggingface.co/zai-org/GLM-5.3","date":"2026-08-14"},{"name":"SWE-Marathon","score":42.5,"metric":"score","harness":"Claude Code 2.1.207","variant":"max effort; modified anti-cheat checks","version":"1.1","source":"https://huggingface.co/zai-org/GLM-5.3"},{"name":"PostTrainBench","score":39.8,"metric":"weighted average","harness":"Claude Code 2.1.207","variant":"max effort; 3 runs; modified external-API checks","source":"https://huggingface.co/zai-org/GLM-5.3"},{"name":"CyberGym","score":84.5,"metric":"pass@1","harness":"Claude Code 2.1.207","variant":"max effort","source":"https://huggingface.co/zai-org/GLM-5.3"},{"name":"ExploitGym","score":105,"metric":"tasks solved","harness":"Claude Code 2.1.207","variant":"max effort; 2h rescaled timeout","dataset":"869 tasks","source":"https://huggingface.co/zai-org/GLM-5.3"},{"name":"ExploitGym","score":130,"metric":"tasks solved","harness":"Claude Code 2.1.207","variant":"max effort; 6h rescaled timeout","dataset":"869 tasks","source":"https://huggingface.co/zai-org/GLM-5.3"},{"name":"ExploitBench","score":54.4,"metric":"average coverage","harness":"Claude Code 2.1.207","variant":"max effort; union over 3 revisions","source":"https://huggingface.co/zai-org/GLM-5.3"},{"name":"Toolathlon-Verified","score":73,"metric":"pass@1","variant":"max effort","source":"https://huggingface.co/zai-org/GLM-5.3"},{"name":"AutomationBench","score":48.2,"metric":"pass@1","variant":"max effort","version":"1.0.6","source":"https://huggingface.co/zai-org/GLM-5.3"},{"name":"Agents' Last Exam","score":28.5,"metric":"score","harness":"Claude Code 2.1.207","variant":"max effort; CLI","source":"https://huggingface.co/zai-org/GLM-5.3"},{"name":"Humanity's Last Exam","score":62.5,"metric":"accuracy","variant":"max effort; with tools","source":"https://huggingface.co/zai-org/GLM-5.3"},{"name":"GDPval-AA","score":1769,"metric":"Elo","variant":"max effort","version":"2","source":"https://huggingface.co/zai-org/GLM-5.3"}]},"zhipuai/glm-4.5":{"id":"zhipuai/glm-4.5","name":"GLM-4.5","description":"Hybrid-reasoning GLM release that made the 4.5 line broadly useful","family":"glm","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/zai-org/GLM-4.5"}],"benchmarks":[{"name":"Artificial Analysis Coding Index","score":26.3,"metric":"index","source":"https://openrouter.ai/z-ai/glm-4.5/benchmarks","date":"2026-03-11"},{"name":"SciCode","score":34.8,"metric":"percent correct","source":"https://openrouter.ai/z-ai/glm-4.5/benchmarks","date":"2026-03-11"},{"name":"Terminal-Bench Hard","score":22,"metric":"success rate","source":"https://openrouter.ai/z-ai/glm-4.5/benchmarks","date":"2026-03-11"}]},"zhipuai/glm-5-turbo":{"id":"zhipuai/glm-5-turbo","name":"GLM-5-Turbo","description":"Faster GLM-5 lane for coding agents that need lower latency","family":"glm","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":131072}},"zhipuai/glm-5.1":{"id":"zhipuai/glm-5.1","name":"GLM-5.1","description":"Strong GLM coding model for agentic engineering, terminals, and repository generation","family":"glm","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-04-07","last_updated":"2026-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/zai-org/GLM-5.1"}],"benchmarks":[{"name":"Artificial Analysis Coding Agent Index","score":52.7,"metric":"average pass@1","harness":"Claude Code","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"SWE-Atlas Codebase QnA","score":73.2,"metric":"pass@1","harness":"Claude Code","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"SWE-Bench Pro","score":19.8,"metric":"pass@1","harness":"Claude Code","dataset":"hard-aa","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"Terminal-Bench","score":65.1,"metric":"pass@1","harness":"Claude Code","version":"2.1","source":"https://artificialanalysis.ai/agents/coding-agents"}]},"zhipuai/glm-4.7-flash":{"id":"zhipuai/glm-4.7-flash","name":"GLM-4.7-Flash","description":"Budget GLM lane for fast coding help, routing, and everyday automation","family":"glm-flash","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-01-19","last_updated":"2026-01-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":200000,"output":131072},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/zai-org/GLM-4.7-Flash"}],"benchmarks":[{"name":"SWE-Bench Verified","score":59.2,"metric":"resolved","source":"https://huggingface.co/zai-org/GLM-4.7-Flash"}]},"zhipuai/glm-5":{"id":"zhipuai/glm-5","name":"GLM-5","description":"General GLM flagship for coding, analysis, and tool-heavy engineering workflows","family":"glm","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/zai-org/GLM-5"}],"benchmarks":[{"name":"SWE-Bench Verified","score":72.8,"metric":"resolved","source":"https://www.swebench.com/"},{"name":"SWE-Atlas Codebase QnA","score":20.5,"metric":"score","harness":"Mini-SWE-Agent","source":"https://labs.scale.com/leaderboard/sweatlas-qna"},{"name":"SWE-Atlas Refactoring","score":24.24,"metric":"score","harness":"Mini-SWE-Agent","source":"https://labs.scale.com/leaderboard/sweatlas-refactoring"},{"name":"SWE-Atlas Test Writing","score":28.74,"metric":"score","harness":"Mini-SWE-Agent","source":"https://labs.scale.com/leaderboard/sweatlas-tw"}]},"zhipuai/glm-5.3-flash":{"id":"zhipuai/glm-5.3-flash","name":"GLM-5.3-Flash","description":"Native multimodal GLM model for efficient coding and long-horizon agent tasks","family":"glm-flash","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/zai-org/GLM-5.3-Flash"}],"benchmarks":[{"name":"Terminal-Bench","score":84.3,"metric":"pass@1","harness":"Claude Code 2.1.207","variant":"max effort","version":"2.1","source":"https://huggingface.co/zai-org/GLM-5.3-Flash"},{"name":"DeepSWE","score":63.4,"metric":"resolved","harness":"mini-swe-agent","variant":"max effort","version":"1.1","source":"https://huggingface.co/zai-org/GLM-5.3-Flash"},{"name":"Agents' Last Exam","score":26.3,"metric":"score","variant":"max effort","source":"https://huggingface.co/zai-org/GLM-5.3-Flash"},{"name":"AutomationBench","score":48.8,"metric":"pass@1","variant":"max effort","version":"1.0.6","source":"https://huggingface.co/zai-org/GLM-5.3-Flash"},{"name":"Humanity's Last Exam","score":55.3,"metric":"accuracy","variant":"max effort; with tools","dataset":"full set","source":"https://huggingface.co/zai-org/GLM-5.3-Flash"},{"name":"GDPval-AA","score":1773,"metric":"Elo","variant":"max effort","version":"2","source":"https://huggingface.co/zai-org/GLM-5.3-Flash"}]},"zhipuai/glm-4.6v-flash":{"id":"zhipuai/glm-4.6v-flash","name":"GLM-4.6V-Flash","description":"Lightweight GLM vision model for visual reasoning, documents, and multimodal agents","family":"glm-flash","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/zai-org/GLM-4.6V-Flash"}]},"zhipuai/glm-image":{"id":"zhipuai/glm-image","type":"image","name":"GLM-Image","description":"GLM-Image is an image generation model adopts a hybrid autoregressive + diffusion decoder architecture. In general image generation quality, GLM‑Image aligns with mainstream latent diffusion approaches, but it shows significant advantages in text-rendering and knowledge‑intensive generation scenarios. It performs especially well in tasks requiring precise semantic understanding and complex information expression, while maintaining strong capabilities in high‑fidelity and fine‑grained detail generation. In addition to text‑to‑image generation, GLM‑Image also supports a rich set of image‑to‑image tasks including image editing, style transfer, identity‑preserving generation, and multi‑subject consistency.","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-01-14","last_updated":"2026-01-14","modalities":{"input":["text","image"],"output":["image"]},"open_weights":true,"limit":{"context":10240,"output":0},"license":"MIT","weights":[{"label":"Hugging Face","url":"https://huggingface.co/zai-org/GLM-Image"}]},"zhipuai/glm-4.6":{"id":"zhipuai/glm-4.6","name":"GLM-4.6","description":"Late GLM-4 workhorse for coding agents, reasoning, and structured tasks","family":"glm","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-30","last_updated":"2025-09-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/zai-org/GLM-4.6"}],"benchmarks":[{"name":"Artificial Analysis Coding Index","score":29.5,"metric":"index","source":"https://openrouter.ai/z-ai/glm-4.6/benchmarks","date":"2026-05-22"},{"name":"SciCode","score":38.4,"metric":"percent correct","source":"https://openrouter.ai/z-ai/glm-4.6/benchmarks","date":"2026-05-22"},{"name":"Terminal-Bench Hard","score":25,"metric":"success rate","source":"https://openrouter.ai/z-ai/glm-4.6/benchmarks","date":"2026-05-22"},{"name":"SWE-Bench Pro","score":9.67,"metric":"resolve rate","dataset":"public","source":"https://labs.scale.com/leaderboard/swe_bench_pro_public"}]},"paddlepaddle/paddleocr-vl":{"id":"paddlepaddle/paddleocr-vl","name":"PaddleOCR VL","description":"Compact vision-language model for multilingual document parsing, OCR, tables, formulas, and charts","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-10-16","last_updated":"2025-10-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"license":"Apache-2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/PaddlePaddle/PaddleOCR-VL"}]},"aleph-alpha/kolibri-1":{"id":"aleph-alpha/kolibri-1","name":"Kolibri 1","description":"Aleph Alpha's German and English mixture-of-experts reasoning model (78B total, 3.46B active) for chat, RAG and agents","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2026-06-18","release_date":"2026-10-03","last_updated":"2026-10-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576},"license":"Apache-2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/Aleph-Alpha/Kolibri-1"}]},"baai/bge-reranker-v2-m3":{"id":"baai/bge-reranker-v2-m3","type":"reranking","name":"BGE Reranker V2 M3","description":"Lightweight multilingual cross-encoder that scores query-document relevance","family":"bge","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-03-15","last_updated":"2024-03-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":0},"license":"Apache-2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/BAAI/bge-reranker-v2-m3"}]},"baai/bge-multilingual-gemma2":{"id":"baai/bge-multilingual-gemma2","type":"embedding","name":"BGE Multilingual Gemma2","description":"Multilingual Gemma 2 based embedding model for retrieval and semantic search","family":"bge","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-07-25","last_updated":"2024-07-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":0},"license":"Gemma","weights":[{"label":"Hugging Face","url":"https://huggingface.co/BAAI/bge-multilingual-gemma2"}]},"baai/bge-m3":{"id":"baai/bge-m3","type":"embedding","name":"BGE M3","description":"Multilingual embedding model combining dense, sparse, and multi-vector retrieval","family":"bge","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-01-30","last_updated":"2024-01-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":0},"license":"MIT","weights":[{"label":"Hugging Face","url":"https://huggingface.co/BAAI/bge-m3"}]},"xai/grok-imagine-image-2.0":{"id":"xai/grok-imagine-image-2.0","type":"image","name":"Grok Imagine Image 2.0","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"grok","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-08-07","last_updated":"2026-08-07","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":64000,"output":0}},"xai/grok-imagine-image":{"id":"xai/grok-imagine-image","type":"image","name":"Grok Imagine Image","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"grok","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-01-28","last_updated":"2026-01-28","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":16000,"output":0}},"xai/grok-tts":{"id":"xai/grok-tts","type":"speech","name":"Grok TTS","description":"Convert text into spoken audio with a single API call. The API supports a rich set of expressive voices, inline speech tags for fine-grained delivery control, and output formats from high-fidelity MP3 to telephony-optimized μ-law.","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":15000,"output":15000}},"xai/grok-4.6":{"id":"xai/grok-4.6","name":"Grok 4.6","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2026-02-01","release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"benchmarks":[{"name":"CursorBench","score":40.4,"metric":"score","variant":"high effort","version":"4.0","source":"https://x.ai/news/grok-4-7","date":"2026-09-21"},{"name":"DeepSWE","score":65.2,"metric":"score","variant":"high effort","version":"1.1","source":"https://x.ai/news/grok-4-7","date":"2026-09-21"},{"name":"EEBench","score":53,"metric":"score","variant":"high effort","source":"https://x.ai/news/grok-4-7","date":"2026-09-21"},{"name":"AA-Briefcase","score":1546,"metric":"Elo","variant":"high effort","version":"1.1","source":"https://x.ai/news/grok-4-7","date":"2026-09-21"},{"name":"Terminal-Bench","score":20.3,"metric":"score","variant":"high effort","version":"4.0","source":"https://x.ai/news/grok-4-7","date":"2026-09-21"},{"name":"Harvey Legal Agent Benchmark","score":15.8,"metric":"score","variant":"high effort","source":"https://x.ai/news/grok-4-7","date":"2026-09-21"},{"name":"HealthBench","score":48.5,"metric":"score","variant":"high effort; professional","source":"https://x.ai/news/grok-4-7","date":"2026-09-21"},{"name":"GDPval","score":1605,"metric":"Elo","variant":"high effort","source":"https://x.ai/news/grok-4-7","date":"2026-09-21"}]},"xai/grok-code-fast-1":{"id":"xai/grok-code-fast-1","name":"x-AI/Grok-Code-Fast 1","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2025-09-02","last_updated":"2025-09-02","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":10000}},"xai/grok-4.3":{"id":"xai/grok-4.3","name":"Grok 4.3","description":"xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000},"benchmarks":[{"name":"Artificial Analysis Intelligence Index","score":53,"metric":"index score","version":"4.0","source":"https://artificialanalysis.ai/articles/xai-launches-grok-4-3-with-improved-agentic-performance-and-lower-pricing","date":"2026-04-30"},{"name":"GDPval-AA","score":1500,"metric":"Elo","source":"https://artificialanalysis.ai/articles/xai-launches-grok-4-3-with-improved-agentic-performance-and-lower-pricing","date":"2026-04-30"},{"name":"τ²-Bench Telecom","score":98,"metric":"success rate","source":"https://artificialanalysis.ai/articles/xai-launches-grok-4-3-with-improved-agentic-performance-and-lower-pricing","date":"2026-04-30"},{"name":"IFBench","score":81,"metric":"accuracy","source":"https://artificialanalysis.ai/articles/xai-launches-grok-4-3-with-improved-agentic-performance-and-lower-pricing","date":"2026-04-30"}]},"xai/grok-4-fast-reasoning":{"id":"xai/grok-4-fast-reasoning","name":"X-Ai/Grok-4-Fast-Reasoning","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2025-12-18","last_updated":"2025-12-18","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":2000000}},"xai/grok-imagine-video-1.5-lite":{"id":"xai/grok-imagine-video-1.5-lite","type":"video","name":"Grok Imagine Video 1.5 Lite","description":"Cost-efficient video model for text-to-video and image-to-video generation","family":"grok","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-01","last_updated":"2026-10-01","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":1024,"output":0}},"xai/grok-4.1-fast-reasoning":{"id":"xai/grok-4.1-fast-reasoning","name":"Grok 4.1 Fast (Reasoning)","description":"xAI's fast agentic tool-calling model with a 2M context window and built-in reasoning","family":"grok","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000}},"xai/grok-4.20-0309-reasoning":{"id":"xai/grok-4.20-0309-reasoning","name":"Grok 4.20 (Reasoning)","description":"Reasoning Grok for document-heavy analysis and long-horizon tool use","family":"grok","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000}},"xai/grok-4.5":{"id":"xai/grok-4.5","name":"Grok 4.5","description":"xAI's Grok model for chat, coding, agentic tools, and lower hallucination risk","family":"grok","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-07-08","last_updated":"2026-07-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"benchmarks":[{"name":"SWE-Bench Pro","score":64.7,"metric":"resolve rate","source":"https://x.ai/news/grok-4-5","date":"2026-07-08"},{"name":"SWE-Bench Multilingual","score":78,"metric":"resolve rate","source":"https://x.ai/news/grok-4-5","date":"2026-07-08"},{"name":"Terminal-Bench","score":83.3,"metric":"success rate","version":"2.1","source":"https://x.ai/news/grok-4-5","date":"2026-07-08"},{"name":"DeepSWE","score":62,"metric":"resolve rate","version":"1.0","source":"https://x.ai/news/grok-4-5","date":"2026-07-08"},{"name":"DeepSWE","score":53,"metric":"resolve rate","harness":"mini-swe-agent","version":"1.1","source":"https://x.ai/news/grok-4-5","date":"2026-07-08"},{"name":"SWE Marathon","score":29,"metric":"pass@1","source":"https://x.ai/news/grok-4-5","date":"2026-07-08"}]},"xai/grok-build-0.1":{"id":"xai/grok-build-0.1","name":"Grok Build 0.1","description":"Fast Grok coding model tuned for agentic engineering and iterative edits","family":"grok-build","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000}},"xai/grok-4-fast-non-reasoning":{"id":"xai/grok-4-fast-non-reasoning","name":"X-Ai/Grok-4-Fast-Non-Reasoning","description":"Fast Grok model for responsive chat, reasoning, and tool-assisted work","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-12-18","last_updated":"2025-12-18","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":2000000}},"xai/grok-imagine-video-1.5":{"id":"xai/grok-imagine-video-1.5","type":"video","name":"Grok Imagine Video 1.5","description":"Video model for text-to-video, image-to-video, and reference-to-video generation with optional voice references","family":"grok","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-05-30","last_updated":"2026-05-30","modalities":{"input":["text","image","audio"],"output":["video"]},"open_weights":false,"limit":{"context":1024,"output":0}},"xai/grok-4.1-fast":{"id":"xai/grok-4.1-fast","name":"Grok 4.1 Fast","description":"xAI's fast agentic tool-calling model with a 2M context window; non-reasoning variant for low-latency responses","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-11-19","last_updated":"2025-11-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":2000000,"output":30000}},"xai/grok-imagine-image-quality":{"id":"xai/grok-imagine-image-quality","type":"image","name":"Grok Imagine Image Quality","description":"Higher-fidelity Grok Imagine image model for prompt-driven generation, editing, and visual design workflows","family":"grok","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-03","last_updated":"2026-04-03","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":16000,"output":0}},"xai/grok-voice-transcribe-2.0":{"id":"xai/grok-voice-transcribe-2.0","type":"transcription","name":"Grok Voice Transcribe 2.0","description":"Speech-to-text model with word-level timestamps, speaker diarization, multichannel audio, and streaming","family":"grok","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-17","last_updated":"2026-09-17","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":0,"output":0}},"xai/grok-4.20-0309-non-reasoning":{"id":"xai/grok-4.20-0309-non-reasoning","name":"Grok 4.20 (Non-Reasoning)","description":"Grok model for agentic tool use, reasoning, coding, and live assistance","family":"grok","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-03-09","last_updated":"2026-03-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":30000}},"xai/grok-4.7":{"id":"xai/grok-4.7","name":"Grok 4.7","description":"xAI's frontier model for long-running agents, coding, knowledge work, and visual projects","family":"grok","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2026-05","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":500000,"output":500000},"benchmarks":[{"name":"CursorBench","score":46.3,"metric":"score","variant":"xhigh effort","version":"4.0","source":"https://x.ai/news/grok-4-7","date":"2026-09-21"},{"name":"DeepSWE","score":71,"metric":"score","variant":"high effort","version":"1.1","source":"https://x.ai/news/grok-4-7","date":"2026-09-21"},{"name":"EEBench","score":64,"metric":"score","variant":"xhigh effort","source":"https://x.ai/news/grok-4-7","date":"2026-09-21"},{"name":"AA-Briefcase","score":1657,"metric":"Elo","variant":"xhigh effort","version":"1.1","source":"https://x.ai/news/grok-4-7","date":"2026-09-21"},{"name":"Terminal-Bench","score":37.6,"metric":"score","variant":"xhigh effort","version":"4.0","source":"https://x.ai/news/grok-4-7","date":"2026-09-21"},{"name":"Harvey Legal Agent Benchmark","score":19.6,"metric":"score","variant":"xhigh effort","source":"https://x.ai/news/grok-4-7","date":"2026-09-21"},{"name":"HealthBench","score":56.7,"metric":"score","variant":"xhigh effort; professional","source":"https://x.ai/news/grok-4-7","date":"2026-09-21"},{"name":"GDPval","score":1695,"metric":"Elo","variant":"xhigh effort","source":"https://x.ai/news/grok-4-7","date":"2026-09-21"}]},"quiverai/arrow-2":{"id":"quiverai/arrow-2","name":"Arrow 2","description":"Fast SVG generation model for creation, vectorization, editing, and animation","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-16","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":65536}},"quiverai/arrow-2-telos":{"id":"quiverai/arrow-2-telos","name":"Arrow 2 Telos","description":"High-fidelity SVG generation model for complex vector work and long-context refinement","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"release_date":"2026-09-16","last_updated":"2026-09-16","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":1050000,"output":65536}},"trendyol/asure-12b":{"id":"trendyol/asure-12b","name":"Trendyol Asure 12B","description":"Turkish-language multimodal instruct model built on Gemma 3 12B for e-commerce text, chat, and image-text tasks","family":"gemma","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-02-19","last_updated":"2026-02-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072},"license":"Gemma","weights":[{"label":"Hugging Face","url":"https://huggingface.co/Trendyol/Trendyol-LLM-Asure-12B"}]},"mindlab/macaron-v1-venti":{"id":"mindlab/macaron-v1-venti","name":"Macaron V1 Venti","description":"GLM-5.2-based Mixture-of-LoRA model with specialists for chat, personal agents, coding, and generative UI","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"license":"MIT","weights":[{"label":"Hugging Face","url":"https://huggingface.co/mindlab-research/Macaron-V1-Venti"}]},"mindlab/macaron-v1-tall":{"id":"mindlab/macaron-v1-tall","name":"Macaron V1 Tall","description":"Qwen3.6-based Mixture-of-LoRA model with specialists for chat, personal agents, coding, and generative UI","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"license":"MIT","weights":[{"label":"Hugging Face","url":"https://huggingface.co/mindlab-research/Macaron-V1-Tall"}]},"aion-labs/aion-2.0":{"id":"aion-labs/aion-2.0","name":"Aion-2.0","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2025-12-21","last_updated":"2025-12-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":32768}},"aion-labs/aion-3.0-mini":{"id":"aion-labs/aion-3.0-mini","name":"Aion-3.0-Mini","description":"Efficient model for low-latency assistance, extraction, and routine automation","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-05-14","last_updated":"2026-05-14","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":32768}},"aion-labs/aion-rp-llama-3.1-8b":{"id":"aion-labs/aion-rp-llama-3.1-8b","name":"Aion-RP 1.0 (8B)","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2023-12-31","release_date":"2025-02-04","last_updated":"2025-02-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":29491}},"aion-labs/aion-3.0":{"id":"aion-labs/aion-3.0","name":"Aion-3.0","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-05-05","last_updated":"2026-05-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":32768}},"cloudflare/clef-flash":{"id":"cloudflare/clef-flash","type":"decision","name":"Clef Flash","description":"Cloudflare's fast 9B open-weight decision model that returns probabilities for typed choice, score and yes/no questions about text or images","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-10-01","last_updated":"2026-10-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":0},"license":"Apache-2.0","links":[{"label":"Announcement","url":"https://blog.cloudflare.com/clef-decision-models/","type":"announcement"},{"label":"Model card","url":"https://huggingface.co/Cloudflare/clef-flash","type":"model_card"}],"weights":[{"label":"Hugging Face","url":"https://huggingface.co/Cloudflare/clef-flash"}]},"cloudflare/clef":{"id":"cloudflare/clef","type":"decision","name":"Clef","description":"Cloudflare's open-weight decision model that returns probabilities for typed choice, score and yes/no questions about text or images","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-10-01","last_updated":"2026-10-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":0},"license":"Apache-2.0","links":[{"label":"Announcement","url":"https://blog.cloudflare.com/clef-decision-models/","type":"announcement"},{"label":"Model card","url":"https://huggingface.co/Cloudflare/clef","type":"model_card"}],"weights":[{"label":"Hugging Face","url":"https://huggingface.co/Cloudflare/clef"}]},"google/gemma-4-E2B-it":{"id":"google/gemma-4-E2B-it","name":"Gemma 4 E2B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/google/gemma-4-E2B-it"}]},"google/gemini-3.1-flash-tts-preview":{"id":"google/gemini-3.1-flash-tts-preview","type":"speech","name":"Gemini 3.1 Flash TTS Preview","description":"Low-latency speech generation with steerable prompts and expressive audio tags","family":"gemini-flash","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-15","last_updated":"2026-04-15","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":8192,"output":16384}},"google/gemini-3.1-pro-preview":{"id":"google/gemini-3.1-pro-preview","name":"Gemini 3.1 Pro Preview","description":"Reasoning-first Gemini preview for agentic coding and complex problem solving","family":"gemini-pro","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"benchmarks":[{"name":"SWE-Bench Pro","score":54.2,"metric":"resolve rate","source":"https://www.anthropic.com/news/claude-opus-4-8","date":"2026-05-28"},{"name":"Terminal-Bench","score":70.3,"metric":"success rate","harness":"Terminus-2","version":"2.1","source":"https://www.anthropic.com/news/claude-opus-4-8","date":"2026-05-28"},{"name":"SWE-Bench Pro","score":46.1,"metric":"resolve rate","dataset":"public","source":"https://labs.scale.com/leaderboard/swe_bench_pro_public"},{"name":"SWE-Atlas Codebase QnA","score":13.5,"metric":"score","harness":"Mini-SWE-Agent","source":"https://labs.scale.com/leaderboard/sweatlas-qna"},{"name":"SWE-Atlas Refactoring","score":33.81,"metric":"score","harness":"Gemini CLI","source":"https://labs.scale.com/leaderboard/sweatlas-refactoring"},{"name":"SWE-Atlas Test Writing","score":29.84,"metric":"score","harness":"Mini-SWE-Agent","source":"https://labs.scale.com/leaderboard/sweatlas-tw"},{"name":"Artificial Analysis Coding Agent Index","score":43,"metric":"average pass@1","harness":"Gemini CLI","variant":"high","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"SWE-Atlas Codebase QnA","score":45.6,"metric":"pass@1","harness":"Gemini CLI","variant":"high","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"SWE-Bench Pro","score":15.1,"metric":"pass@1","harness":"Gemini CLI","variant":"high","dataset":"hard-aa","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"Terminal-Bench","score":68.3,"metric":"pass@1","harness":"Gemini CLI","variant":"high","version":"2.1","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"GPQA Diamond","score":94.3,"metric":"accuracy","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"Humanity's Last Exam","score":44.4,"metric":"accuracy","dataset":"full set, text + MM","source":"https://deepmind.google/models/gemini/flash/","date":"2026-05-19"},{"name":"ARC-AGI-2","score":77.1,"metric":"accuracy","source":"https://deepmind.google/models/gemini/flash/","date":"2026-05-19"},{"name":"MMMU Pro","score":80.5,"metric":"accuracy","variant":"no tools","source":"https://deepmind.google/models/gemini/flash/","date":"2026-05-19"},{"name":"MCP Atlas","score":78.2,"metric":"success rate","source":"https://deepmind.google/models/gemini/flash/","date":"2026-05-19"},{"name":"OSWorld-Verified","score":76.2,"metric":"success rate","source":"https://deepmind.google/models/gemini/flash/","date":"2026-05-19"},{"name":"CharXiv Reasoning","score":83.3,"metric":"accuracy","variant":"no tools","source":"https://deepmind.google/models/gemini/flash/","date":"2026-05-19"},{"name":"GDPval-AA","score":1314,"metric":"Elo","source":"https://deepmind.google/models/gemini/flash/","date":"2026-05-19"}]},"google/gemma-3-12b-it":{"id":"google/gemma-3-12b-it","name":"Gemma 3 12B IT","description":"Open multimodal Gemma instruction model for multilingual text generation and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/google/gemma-3-12b-it"}]},"google/gemini-3-pro-image":{"id":"google/gemini-3-pro-image","type":"image","name":"Nano Banana Pro","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":true,"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":32768}},"google/gemini-3.1-flash-image":{"id":"google/gemini-3.1-flash-image","type":"image","name":"Nano Banana 2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","video","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":32768}},"google/gemini-2.5-flash-image":{"id":"google/gemini-2.5-flash-image","type":"image","name":"Nano Banana","description":"Nano Banana image model for fast generation, edits, and character-consistent assets","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":32768,"output":32768}},"google/gemini-3.5-live-translate-preview":{"id":"google/gemini-3.5-live-translate-preview","type":"realtime","name":"Gemini 3.5 Live Translate Preview","description":"Low-latency audio-to-audio model for real-time speech translation across 70+ languages","family":"gemini-pro","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["audio"],"output":["audio","text"]},"open_weights":false,"limit":{"context":131072,"output":65536}},"google/gemma-3n-e2b-it":{"id":"google/gemma-3n-e2b-it","name":"Gemma 3n E2B IT","description":"Efficient on-device Gemma 3n model for text, image, and audio understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-06-26","last_updated":"2025-06-26","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"license":"Gemma Terms of Use","weights":[{"label":"Hugging Face","url":"https://huggingface.co/google/gemma-3n-E2B-it"}]},"google/veo-3.1-lite-generate-preview":{"id":"google/veo-3.1-lite-generate-preview","type":"video","name":"Veo 3.1 Lite Preview","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"veo","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-03-31","last_updated":"2026-03-31","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":1024,"output":0}},"google/veo-3.1-generate-preview":{"id":"google/veo-3.1-generate-preview","type":"video","name":"Veo 3.1 Preview","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"veo","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-10-15","last_updated":"2026-01","modalities":{"input":["text","image"],"output":["video"]},"open_weights":false,"limit":{"context":1024,"output":1}},"google/gemini-3.8-flash":{"id":"google/gemini-3.8-flash","name":"Gemini 3.8 Flash","description":"Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows","family":"gemini-flash","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"benchmarks":[{"name":"DeepSWE","score":73.7,"metric":"pass@1","harness":"mini-swe-agent","variant":"high thinking","version":"1.1","source":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/3-8-flash-and-3-8-flash-cyber/"},{"name":"GDPval-AA","score":1545,"metric":"Elo","version":"2","source":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/3-8-flash-and-3-8-flash-cyber/"},{"name":"Finance Agent","score":61.4,"metric":"pass@1","version":"2","source":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/3-8-flash-and-3-8-flash-cyber/"},{"name":"Harvey Legal Agent Benchmark","score":10,"metric":"all-pass rate","source":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/3-8-flash-and-3-8-flash-cyber/"},{"name":"Terminal-Bench","score":89.4,"metric":"pass@1","harness":"Terminus 2","version":"2.1","source":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/3-8-flash-and-3-8-flash-cyber/"},{"name":"Terminal-Bench","score":19.1,"metric":"pass@1","version":"4.0","source":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/3-8-flash-and-3-8-flash-cyber/"},{"name":"GDP.PDF","score":35,"metric":"all-pass rate","source":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/3-8-flash-and-3-8-flash-cyber/"},{"name":"CharXiv Reasoning","score":86.2,"metric":"pass@1","variant":"without tools","source":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/3-8-flash-and-3-8-flash-cyber/"},{"name":"LVBench","score":87.8,"metric":"pass@1","variant":"agentic","source":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/3-8-flash-and-3-8-flash-cyber/"},{"name":"LVBench","score":87.1,"metric":"pass@1","variant":"static; 1024 frames; without tools","source":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/3-8-flash-and-3-8-flash-cyber/"},{"name":"HLE-Verified","score":54.9,"metric":"accuracy","dataset":"1811 verified and revised items","source":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/3-8-flash-and-3-8-flash-cyber/"},{"name":"OSWorld","score":59,"metric":"partial score","harness":"Gemini CUA","variant":"batched tool calls; best of 3 runs; 500 steps","dataset":"before 2026-08-08 patch","version":"2.0","source":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/3-8-flash-and-3-8-flash-cyber/"},{"name":"BioMysteryBench","score":88.8,"metric":"pass@1","variant":"with tools","dataset":"human-solvable","source":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/3-8-flash-and-3-8-flash-cyber/"},{"name":"BioMysteryBench","score":56.5,"metric":"pass@1","variant":"with tools","dataset":"human-difficult","source":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/3-8-flash-and-3-8-flash-cyber/"},{"name":"LAB-Bench","score":86.2,"metric":"macro-average","variant":"with tools","dataset":"11 subtasks","version":"2","source":"https://blog.google/innovation-and-ai/models-and-research/gemini-models/3-8-flash-and-3-8-flash-cyber/"}]},"google/gemma-3-4b-it":{"id":"google/gemma-3-4b-it","name":"Gemma 3 4B IT","description":"Open multimodal Gemma instruction model for efficient text generation and image understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/google/gemma-3-4b-it"}]},"google/lyria-3-clip-preview":{"id":"google/lyria-3-clip-preview","name":"Lyria 3 Clip Preview","description":"Music generation model for short 30-second clips, loops, and previews from text or image prompts","family":"lyria","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-03-25","last_updated":"2026-03-25","modalities":{"input":["text","image"],"output":["text","audio"]},"open_weights":false,"limit":{"context":131072,"output":65536}},"google/lyria-3-pro-preview":{"id":"google/lyria-3-pro-preview","name":"Lyria 3 Pro Preview","description":"Music generation model for full-length songs from text or images with vocals and structure","family":"lyria","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-03-25","last_updated":"2026-03-25","modalities":{"input":["text","image"],"output":["text","audio"]},"open_weights":false,"limit":{"context":131072,"output":8192}},"google/gemini-flash-lite-latest":{"id":"google/gemini-flash-lite-latest","name":"Gemini Flash-Lite Latest","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536}},"google/gemini-3.7-flash":{"id":"google/gemini-3.7-flash","name":"Gemini 3.7 Flash","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"benchmarks":[{"name":"FrontierCode","score":43.6,"metric":"score","version":"1.1 Main","source":"https://deepmind.google/models/model-cards/gemini-3-7-flash/","date":"2026-08-13"},{"name":"DeepSWE","score":65.3,"metric":"resolve rate","version":"1.1","source":"https://deepmind.google/models/model-cards/gemini-3-7-flash/","date":"2026-08-13"},{"name":"Terminal-Bench","score":85.8,"metric":"accuracy","version":"2.1","source":"https://deepmind.google/models/model-cards/gemini-3-7-flash/","date":"2026-08-13"},{"name":"AutomationBench","score":30.4,"metric":"accuracy","dataset":"private set","source":"https://deepmind.google/models/model-cards/gemini-3-7-flash/","date":"2026-08-13"},{"name":"GDP.pdf","score":34,"metric":"accuracy","source":"https://deepmind.google/models/model-cards/gemini-3-7-flash/","date":"2026-08-13"},{"name":"GDM-MRCR","score":97,"metric":"accuracy","variant":"128k average, 8-needle","version":"v2","source":"https://deepmind.google/models/model-cards/gemini-3-7-flash/","date":"2026-08-13"}]},"google/gemini-omni-1.1-flash":{"id":"google/gemini-omni-1.1-flash","type":"video","name":"Gemini Omni 1.1 Flash","description":"Video generation and editing model for scene extension, frame interpolation, and reference-guided multimodal control","family":"gemini","attachment":true,"reasoning":true,"tool_call":false,"temperature":false,"release_date":"2026-08-27","last_updated":"2026-08-27","modalities":{"input":["text","image","video"],"output":["video"]},"open_weights":false,"limit":{"context":1048576,"output":57920}},"google/gemini-robotics-er-1.6-preview":{"id":"google/gemini-robotics-er-1.6-preview","name":"Gemini Robotics-ER 1.6 Preview","description":"Vision-language model for embodied reasoning: spatial understanding, task planning, and physical-world agentic robotics","family":"gemini","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-14","last_updated":"2026-04-14","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":65536}},"google/gemini-2.5-flash-preview-05-20":{"id":"google/gemini-2.5-flash-preview-05-20","name":"Gemini 2.5 Flash Preview 05-20","description":"Preview snapshot of Gemini 2.5 Flash with controllable thinking budgets","family":"gemini-flash","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-05-20","last_updated":"2025-05-20","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536}},"google/gemma-2-2b-it":{"id":"google/gemma-2-2b-it","name":"Gemma 2 2B IT","description":"Small open Gemma 2 instruction model for lightweight chat and on-device use","family":"gemma","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2024-07-31","last_updated":"2024-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":8192},"license":"Gemma Terms of Use","weights":[{"label":"Hugging Face","url":"https://huggingface.co/google/gemma-2-2b-it"}]},"google/gemini-2.0-flash-lite":{"id":"google/gemini-2.0-flash-lite","name":"Gemini 2.0 Flash-Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2024-12-11","last_updated":"2024-12-11","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":8192}},"google/gemini-3-pro-preview":{"id":"google/gemini-3-pro-preview","name":"Gemini 3 Pro Preview","description":"Preview Gemini flagship for complex reasoning, coding, and rich multimodal prompts","family":"gemini-pro","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-11-18","last_updated":"2025-11-18","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"benchmarks":[{"name":"SWE-Bench Pro","score":43.3,"metric":"resolve rate","dataset":"public","source":"https://labs.scale.com/leaderboard/swe_bench_pro_public"}]},"google/gemma-4-31b-it":{"id":"google/gemma-4-31b-it","name":"Gemma 4 31B IT","description":"Largest Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/google/gemma-4-31B-it"}]},"google/gemini-2.5-flash-lite-preview-06-17":{"id":"google/gemini-2.5-flash-lite-preview-06-17","name":"Gemini 2.5 Flash-Lite Preview 06-17","description":"First preview of Gemini 2.5 Flash-Lite, the low-cost, low-latency 2.5 model","family":"gemini-flash-lite","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536}},"google/gemini-3.5-flash":{"id":"google/gemini-3.5-flash","name":"Gemini 3.5 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"benchmarks":[{"name":"Terminal-Bench","score":76.2,"metric":"success rate","harness":"Terminus-2","version":"2.1","source":"https://deepmind.google/models/gemini/flash/","date":"2026-05-19"},{"name":"SWE-Bench Pro","score":55.1,"metric":"resolve rate","variant":"single attempt","dataset":"public","source":"https://deepmind.google/models/gemini/flash/","date":"2026-05-19"},{"name":"MCP Atlas","score":83.6,"metric":"success rate","source":"https://deepmind.google/models/gemini/flash/","date":"2026-05-19"},{"name":"Toolathlon","score":56.5,"metric":"success rate","source":"https://deepmind.google/models/gemini/flash/","date":"2026-05-19"},{"name":"OSWorld-Verified","score":78.4,"metric":"success rate","source":"https://deepmind.google/models/gemini/flash/","date":"2026-05-19"},{"name":"MMMU Pro","score":83.6,"metric":"accuracy","variant":"no tools","source":"https://deepmind.google/models/gemini/flash/","date":"2026-05-19"},{"name":"CharXiv Reasoning","score":84.2,"metric":"accuracy","variant":"no tools","source":"https://deepmind.google/models/gemini/flash/","date":"2026-05-19"},{"name":"Humanity's Last Exam","score":40.2,"metric":"accuracy","dataset":"full set, text + MM","source":"https://deepmind.google/models/gemini/flash/","date":"2026-05-19"},{"name":"ARC-AGI-2","score":72.1,"metric":"accuracy","source":"https://deepmind.google/models/gemini/flash/","date":"2026-05-19"},{"name":"GDPval-AA","score":1656,"metric":"Elo","source":"https://deepmind.google/models/gemini/flash/","date":"2026-05-19"}]},"google/gemini-2.5-computer-use-preview-10-2025":{"id":"google/gemini-2.5-computer-use-preview-10-2025","name":"Gemini 2.5 Computer Use Preview","description":"Specialized Gemini 2.5 model for browser-control agents that automate UI tasks","family":"gemini-pro","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-10-07","last_updated":"2025-10-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":64000}},"google/gemini-omni-flash-preview":{"id":"google/gemini-omni-flash-preview","type":"video","name":"Gemini Omni Flash Preview","description":"Video generation and editing model for fast, conversational text- and image-to-video workflows","family":"gemini","attachment":true,"reasoning":true,"tool_call":false,"temperature":false,"release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","video"],"output":["video"]},"open_weights":false,"limit":{"context":1048576,"output":57920},"benchmarks":[{"name":"LMArena Text-to-Video Arena","score":1527,"metric":"Elo","source":"https://venturebeat.com/technology/googles-gemini-omni-flash-hits-the-api-turning-enterprise-video-production-into-a-conversation","date":"2026-06-30"}]},"google/diffusiongemma-26b-a4b-it":{"id":"google/diffusiongemma-26b-a4b-it","name":"DiffusionGemma 26B A4B IT","description":"Gemma 4 based discrete text diffusion model for fast chat, reasoning, and tool use","family":"gemma","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-09","last_updated":"2026-07-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"license":"Apache 2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/google/diffusiongemma-26B-A4B-it"}]},"google/veo-3.1-fast-generate-preview":{"id":"google/veo-3.1-fast-generate-preview","type":"video","name":"Veo 3.1 Fast Preview","description":"Video model for prompt-guided generation, editing, and motion workflows","family":"veo","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-10-15","last_updated":"2026-01-01","modalities":{"input":["text","image","video"],"output":["video"]},"open_weights":false,"limit":{"context":1024,"output":0}},"google/gemini-2.5-pro-tts":{"id":"google/gemini-2.5-pro-tts","type":"speech","name":"Gemini 2.5 Pro TTS","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"gemini-pro","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2025-01","release_date":"2025-09-30","last_updated":"2025-12-10","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":32768,"output":16384}},"google/gemini-3-pro-image-preview":{"id":"google/gemini-3-pro-image-preview","type":"image","name":"Nano Banana Pro Preview","description":"Nano Banana Pro for higher-fidelity image generation and design-heavy edits","family":"gemini-pro","attachment":true,"reasoning":true,"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-11-20","last_updated":"2025-11-20","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":32768}},"google/gemini-2.5-flash-lite":{"id":"google/gemini-2.5-flash-lite","name":"Gemini 2.5 Flash-Lite","description":"Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents","family":"gemini-flash-lite","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"benchmarks":[{"name":"Artificial Analysis Coding Index","score":9.5,"metric":"index","source":"https://openrouter.ai/google/gemini-2.5-flash-lite/benchmarks","date":"2026-03-11"},{"name":"SciCode","score":19.3,"metric":"percent correct","source":"https://openrouter.ai/google/gemini-2.5-flash-lite/benchmarks","date":"2026-03-11"},{"name":"Terminal-Bench Hard","score":4.5,"metric":"success rate","source":"https://openrouter.ai/google/gemini-2.5-flash-lite/benchmarks","date":"2026-03-11"}]},"google/gemma-4-E4B-it":{"id":"google/gemma-4-E4B-it","name":"Gemma 4 E4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/google/gemma-4-E4B-it"}]},"google/gemini-3.1-flash-lite-preview":{"id":"google/gemini-3.1-flash-lite-preview","name":"Gemini 3.1 Flash Lite Preview","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536}},"google/gemini-3.1-pro-preview-customtools":{"id":"google/gemini-3.1-pro-preview-customtools","name":"Gemini 3.1 Pro Preview Custom Tools","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-19","last_updated":"2026-02-19","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536}},"google/gemini-3-flash-preview":{"id":"google/gemini-3-flash-preview","name":"Gemini 3 Flash Preview","description":"New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs","family":"gemini-flash","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-12-17","last_updated":"2025-12-17","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"benchmarks":[{"name":"SWE-Bench Pro","score":34.63,"metric":"resolve rate","dataset":"public","source":"https://labs.scale.com/leaderboard/swe_bench_pro_public"},{"name":"SWE-Atlas Codebase QnA","score":8.2,"metric":"score","harness":"Mini-SWE-Agent","source":"https://labs.scale.com/leaderboard/sweatlas-qna"},{"name":"SWE-Atlas Refactoring","score":10,"metric":"score","harness":"Mini-SWE-Agent","source":"https://labs.scale.com/leaderboard/sweatlas-refactoring"},{"name":"SWE-Atlas Test Writing","score":30.3,"metric":"score","harness":"Mini-SWE-Agent","source":"https://labs.scale.com/leaderboard/sweatlas-tw"}]},"google/gemini-nano-banana-2.1":{"id":"google/gemini-nano-banana-2.1","type":"image","name":"Nano Banana 2.1","description":"Image model for prompt-driven generation, multi-turn editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"tool_call":false,"temperature":false,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image","video"],"output":["text","image"]},"open_weights":false,"limit":{"context":131072,"output":32768}},"google/gemini-2.5-flash":{"id":"google/gemini-2.5-flash","name":"Gemini 2.5 Flash","description":"Fast Gemini workhorse for multimodal apps where latency and price matter","family":"gemini-flash","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"benchmarks":[{"name":"Aider Polyglot","score":55.1,"metric":"percent correct","source":"https://aider.chat/docs/leaderboards/","date":"2025-05-25"},{"name":"Artificial Analysis Coding Index","score":22.2,"metric":"index","source":"https://openrouter.ai/google/gemini-2.5-flash/benchmarks","date":"2026-06-02"},{"name":"SciCode","score":39.4,"metric":"percent correct","source":"https://openrouter.ai/google/gemini-2.5-flash/benchmarks","date":"2026-06-02"},{"name":"Terminal-Bench Hard","score":13.6,"metric":"success rate","source":"https://openrouter.ai/google/gemini-2.5-flash/benchmarks","date":"2026-06-02"}]},"google/gemini-3.5-transcribe-live":{"id":"google/gemini-3.5-transcribe-live","type":"transcription","name":"Gemini 3.5 Transcribe Live","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"gemini","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":0,"output":0}},"google/gemini-flash-latest":{"id":"google/gemini-flash-latest","name":"Gemini Flash Latest","description":"High-efficiency Gemini model for agentic workflows, coding, and multimodal reasoning","family":"gemini-flash","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536}},"google/deep-research-max-preview-04-2026":{"id":"google/deep-research-max-preview-04-2026","name":"Deep Research Max Preview","description":"Maximum-comprehensiveness agentic researcher for multi-step investigation, synthesis, and cited reports","family":"gemini-pro","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":1048576,"output":65536}},"google/gemma-2-27b-it":{"id":"google/gemma-2-27b-it","name":"Gemma 2 27B","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-06-30","release_date":"2024-07-13","last_updated":"2024-07-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":2048}},"google/gemini-3.6-flash":{"id":"google/gemini-3.6-flash","name":"Gemini 3.6 Flash","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"benchmarks":[{"name":"SWE-Bench Pro","score":58.7,"metric":"resolve rate","harness":"Antigravity","source":"https://deepmind.google/models/evals-methodology/gemini-3-6-flash/","date":"2026-07-21"},{"name":"DeepSWE","score":49,"metric":"resolve rate","variant":"high reasoning","version":"1.1","source":"https://deepmind.google/models/evals-methodology/gemini-3-6-flash/","date":"2026-07-21"},{"name":"Terminal-Bench","score":78,"metric":"accuracy","harness":"Terminus 2","version":"2.1","source":"https://deepmind.google/models/evals-methodology/gemini-3-6-flash/","date":"2026-07-21"},{"name":"MLE-Bench","score":63.9,"metric":"average position score","dataset":"Partial 30","source":"https://deepmind.google/models/evals-methodology/gemini-3-6-flash/","date":"2026-07-21"},{"name":"GDPval-AA","score":1421,"metric":"Elo","version":"v2","source":"https://deepmind.google/models/evals-methodology/gemini-3-6-flash/","date":"2026-07-21"},{"name":"OSWorld-Verified","score":83,"metric":"success rate","source":"https://deepmind.google/models/evals-methodology/gemini-3-6-flash/","date":"2026-07-21"},{"name":"CharXiv Reasoning","score":85.2,"metric":"accuracy","variant":"no tools","source":"https://deepmind.google/models/evals-methodology/gemini-3-6-flash/","date":"2026-07-21"},{"name":"CharXiv Reasoning","score":89.4,"metric":"accuracy","variant":"with tools","source":"https://deepmind.google/models/evals-methodology/gemini-3-6-flash/","date":"2026-07-21"},{"name":"GDM-MRCR","score":91.8,"metric":"accuracy","variant":"128k average, 8-needle","version":"v2","source":"https://deepmind.google/models/evals-methodology/gemini-3-6-flash/","date":"2026-07-21"},{"name":"GDM-MRCR","score":54,"metric":"accuracy","variant":"1M pointwise, 8-needle","version":"v2","source":"https://deepmind.google/models/evals-methodology/gemini-3-6-flash/","date":"2026-07-21"}]},"google/gemini-embedding-001":{"id":"google/gemini-embedding-001","type":"embedding","name":"Gemini Embedding 001","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"gemini","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2025-05","release_date":"2025-05-20","last_updated":"2025-05-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":2048,"output":1}},"google/gemini-3.1-flash-lite":{"id":"google/gemini-3.1-flash-lite","name":"Gemini 3.1 Flash Lite","description":"Low-latency Gemini model for high-volume multimodal and agent workloads","family":"gemini-flash-lite","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536}},"google/gemini-3.1-flash-lite-image":{"id":"google/gemini-3.1-flash-lite-image","type":"image","name":"Nano Banana 2 Lite","description":"Fastest, most cost-efficient Gemini image model for high-volume 1K generation and editing","family":"gemini-flash-lite","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":4096}},"google/gemini-embedding-2":{"id":"google/gemini-embedding-2","type":"embedding","name":"Gemini Embedding 2","description":"Multimodal embedding model mapping text, images, video, audio, and PDFs into a unified embedding space","family":"gemini","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"knowledge":"2025-11","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":3072}},"google/gemini-2.0-flash":{"id":"google/gemini-2.0-flash","name":"Gemini 2.0 Flash","description":"Earlier Gemini Flash workhorse for responsive multimodal apps and tool use","family":"gemini-flash","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2024-12-11","last_updated":"2024-12-11","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":8192}},"google/gemini-2.5-flash-lite-preview-09-2025":{"id":"google/gemini-2.5-flash-lite-preview-09-2025","name":"Gemini 2.5 Flash-Lite Preview 09-2025","description":"September 2025 preview of Gemini 2.5 Flash-Lite with better instruction following and multimodal quality","family":"gemini-flash-lite","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-09-25","last_updated":"2025-09-25","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536}},"google/gemma-4-12b-it":{"id":"google/gemma-4-12b-it","name":"Gemma 4 12B IT","description":"Compact Gemma 4 instruction model for open, self-hosted chat and reasoning","family":"gemma","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/google/gemma-4-12B-it"}]},"google/gemini-2.5-pro":{"id":"google/gemini-2.5-pro","name":"Gemini 2.5 Pro","description":"Google's proven reasoning model for coding, math, and multimodal analysis","family":"gemini-pro","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"benchmarks":[{"name":"Aider Polyglot","score":83.1,"metric":"percent correct","source":"https://aider.chat/docs/leaderboards/","date":"2025-06-06"},{"name":"Artificial Analysis Coding Index","score":32,"metric":"index","source":"https://openrouter.ai/google/gemini-2.5-pro/benchmarks","date":"2026-06-02"},{"name":"SciCode","score":42.8,"metric":"percent correct","source":"https://openrouter.ai/google/gemini-2.5-pro/benchmarks","date":"2026-06-02"},{"name":"Terminal-Bench Hard","score":26.5,"metric":"success rate","source":"https://openrouter.ai/google/gemini-2.5-pro/benchmarks","date":"2026-06-02"}]},"google/gemini-3.1-flash-live-preview":{"id":"google/gemini-3.1-flash-live-preview","type":"realtime","name":"Gemini 3.1 Flash Live Preview","description":"High-quality, low-latency Live API model for real-time dialogue and voice-first AI applications","family":"gemini-flash","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-03-26","last_updated":"2026-03-26","modalities":{"input":["text","image","video","audio"],"output":["text","audio"]},"open_weights":false,"limit":{"context":131072,"output":65536}},"google/gemma-4-26b-a4b-it":{"id":"google/gemma-4-26b-a4b-it","name":"Gemma 4 26B A4B IT","description":"Open Gemma instruction model for efficient chat and self-hosted deployments","family":"gemma","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/google/gemma-4-26B-A4B-it"}]},"google/gemini-4-argon":{"id":"google/gemini-4-argon","name":"Gemini 4 Argon","description":"Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents","family":"gemini","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2026-03","release_date":"2026-10-13","last_updated":"2026-10-13","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536}},"google/gemini-2.5-pro-preview":{"id":"google/gemini-2.5-pro-preview","name":"Gemini 2.5 Pro Preview 06-05","description":"Advanced Gemini model for complex reasoning, coding, and multimodal analysis","family":"gemini","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-01-31","release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["pdf","image","text","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536}},"google/gemini-3.5-flash-lite":{"id":"google/gemini-3.5-flash-lite","name":"Gemini 3.5 Flash Lite","description":"Fast Gemini model balancing multimodal reasoning, tool use, and cost","family":"gemini-flash-lite","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2026-03","release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536},"benchmarks":[{"name":"SWE-Bench Pro","score":54.2,"metric":"resolve rate","source":"https://deepmind.google/models/model-cards/gemini-3-5-flash-lite/","date":"2026-07-21"},{"name":"Terminal-Bench","score":54,"metric":"accuracy","harness":"Terminus 2","version":"2.1","source":"https://deepmind.google/models/model-cards/gemini-3-5-flash-lite/","date":"2026-07-21"},{"name":"MLE-Bench","score":39.2,"metric":"average position score","source":"https://deepmind.google/models/model-cards/gemini-3-5-flash-lite/","date":"2026-07-21"},{"name":"GDPval-AA","score":1140,"metric":"Elo","version":"v2","source":"https://deepmind.google/models/model-cards/gemini-3-5-flash-lite/","date":"2026-07-21"},{"name":"OSWorld-Verified","score":74,"metric":"success rate","source":"https://deepmind.google/models/model-cards/gemini-3-5-flash-lite/","date":"2026-07-21"},{"name":"CharXiv Reasoning","score":74.5,"metric":"accuracy","variant":"no tools","source":"https://deepmind.google/models/model-cards/gemini-3-5-flash-lite/","date":"2026-07-21"},{"name":"CharXiv Reasoning","score":76.5,"metric":"accuracy","variant":"with tools","source":"https://deepmind.google/models/model-cards/gemini-3-5-flash-lite/","date":"2026-07-21"},{"name":"GDM-MRCR","score":72.2,"metric":"accuracy","variant":"128k average, 8-needle","version":"v2","source":"https://deepmind.google/models/model-cards/gemini-3-5-flash-lite/","date":"2026-07-21"},{"name":"GDM-MRCR","score":21.3,"metric":"accuracy","variant":"1M pointwise, 8-needle","version":"v2","source":"https://deepmind.google/models/model-cards/gemini-3-5-flash-lite/","date":"2026-07-21"}]},"google/gemma-3n-e4b-it":{"id":"google/gemma-3n-e4b-it","name":"Gemma 3n E4B IT","description":"Efficient on-device Gemma 3n model for text, image, and audio understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-06","release_date":"2025-06-26","last_updated":"2025-06-26","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"license":"Gemma Terms of Use","weights":[{"label":"Hugging Face","url":"https://huggingface.co/google/gemma-3n-E4B-it"}]},"google/gemini-2.5-flash-tts":{"id":"google/gemini-2.5-flash-tts","type":"speech","name":"Gemini 2.5 Flash TTS","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"gemini-flash","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2025-09-30","last_updated":"2025-12-10","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":32768,"output":16384}},"google/deep-research-preview-04-2026":{"id":"google/deep-research-preview-04-2026","name":"Gemini Deep Research Preview","description":"Agentic model for autonomous multi-step research, synthesis, and cited reports","family":"gemini-pro","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text","image","video","audio","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":1048576,"output":65536}},"google/gemma-3-27b-it":{"id":"google/gemma-3-27b-it","name":"Gemma 3 27B IT","description":"Largest open Gemma 3 instruction model for multilingual text generation and visual understanding","family":"gemma","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-03-12","last_updated":"2025-03-12","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/google/gemma-3-27b-it"}]},"google/gemini-3.1-flash-image-preview":{"id":"google/gemini-3.1-flash-image-preview","type":"image","name":"Nano Banana 2 Preview","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gemini-flash","attachment":true,"reasoning":true,"tool_call":false,"temperature":true,"knowledge":"2025-01","release_date":"2026-02-26","last_updated":"2026-02-26","modalities":{"input":["text","image","pdf"],"output":["text","image"]},"open_weights":false,"limit":{"context":65536,"output":65536}},"undi95/remm-slerp-l2-13b":{"id":"undi95/remm-slerp-l2-13b","name":"ReMM SLERP 13B","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2023-06-30","release_date":"2023-07-22","last_updated":"2023-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":6144,"output":5529}},"essentialai/rnj-1-instruct":{"id":"essentialai/rnj-1-instruct","name":"Rnj-1 Instruct","description":"Open 8B Essential AI model for code, STEM reasoning, and agentic tool use","family":"rnj","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-12-04","last_updated":"2025-12-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"license":"Apache 2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/EssentialAI/rnj-1-instruct"}]},"inclusionai/ling-3.0-flash-fin":{"id":"inclusionai/ling-3.0-flash-fin","name":"Ling 3.0 Flash Fin","description":"Finance-enhanced model for financial research, multi-step investment workflows, and long-horizon planning and execution","family":"ling","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-08-27","last_updated":"2026-09-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"license":"MIT","weights":[{"label":"Hugging Face","url":"https://huggingface.co/inclusionAI/Ling-3.0-flash-Fin"}],"benchmarks":[{"name":"FinFIRST","score":52.85,"metric":"score","harness":"ReAct with web search, visit, and Python","variant":"highest reasoning effort","source":"https://huggingface.co/inclusionAI/Ling-3.0-flash-Fin"},{"name":"FinSearchComp Verified","score":77.04,"metric":"score","harness":"ReAct with web search, visit, and Python","variant":"highest reasoning effort","dataset":"expert-validated internal set","source":"https://huggingface.co/inclusionAI/Ling-3.0-flash-Fin"},{"name":"FinCRAFT","score":51.74,"metric":"score","variant":"highest reasoning effort","dataset":"internal","source":"https://huggingface.co/inclusionAI/Ling-3.0-flash-Fin"},{"name":"Finance Agent","score":69.19,"metric":"strict-pass","variant":"highest reasoning effort; 10 runs","version":"1.1","source":"https://huggingface.co/inclusionAI/Ling-3.0-flash-Fin"},{"name":"Finance Agent","score":59.81,"metric":"strict-pass","variant":"highest reasoning effort; 20 runs","version":"2","source":"https://huggingface.co/inclusionAI/Ling-3.0-flash-Fin"},{"name":"APEX-Agents","score":29.17,"metric":"score","variant":"highest reasoning effort","source":"https://huggingface.co/inclusionAI/Ling-3.0-flash-Fin"},{"name":"SpreadsheetBench","score":86.5,"metric":"score","harness":"Claude Code 2.1.173","variant":"highest reasoning effort; temperature=0.6; 120 turns; 3h timeout","version":"1","source":"https://huggingface.co/inclusionAI/Ling-3.0-flash-Fin"},{"name":"SpreadsheetBench","score":21.81,"metric":"score","harness":"Claude Code 2.1.173","variant":"highest reasoning effort; temperature=0.6; 300 turns; 3h timeout","version":"2","source":"https://huggingface.co/inclusionAI/Ling-3.0-flash-Fin"},{"name":"tau3-bench","score":41,"metric":"score","variant":"highest reasoning effort; banking","source":"https://huggingface.co/inclusionAI/Ling-3.0-flash-Fin"}]},"inclusionai/ling-3.0-flash-sante":{"id":"inclusionai/ling-3.0-flash-sante","name":"Ling 3.0 Flash Sante","description":"Health and medicine specialization of Ling 3.0 Flash for medical reasoning and evidence-based retrieval","family":"ling","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768}},"inclusionai/ling-3.1-flash":{"id":"inclusionai/ling-3.1-flash","name":"Ling 3.1 Flash","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"ling","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768}},"inclusionai/ling-3.0-flash":{"id":"inclusionai/ling-3.0-flash","name":"Ling 3.0 Flash","description":"Hybrid-linear MoE reasoning model with 124B total and 5.1B active parameters for efficient agents","family":"ling","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-07-23","last_updated":"2026-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"license":"MIT","weights":[{"label":"Hugging Face","url":"https://huggingface.co/inclusionAI/Ling-3.0-flash"}]},"inclusionai/ling-3.0-flash-vl":{"id":"inclusionai/ling-3.0-flash-vl","name":"Ling 3.0 Flash VL","description":"Native multimodal hybrid MoE model for image and video understanding, reasoning, and visual agents","family":"ling","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"license":"MIT","weights":[{"label":"Hugging Face","url":"https://huggingface.co/inclusionAI/Ling-3.0-flash-VL"}]},"microsoft/phi-4-multimodal-instruct":{"id":"microsoft/phi-4-multimodal-instruct","name":"Phi-4-multimodal-instruct","description":"Compact Microsoft multimodal model for text, image, and speech understanding","family":"phi","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-02-26","last_updated":"2025-02-26","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"license":"MIT","weights":[{"label":"Hugging Face","url":"https://huggingface.co/microsoft/Phi-4-multimodal-instruct"}]},"microsoft/mai-code-1.1-flash":{"id":"microsoft/mai-code-1.1-flash","name":"MAI-Code-1.1-Flash","description":"Microsoft coding model with native vision support, optimized for fast and efficient software development","family":"mai","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"links":[{"label":"Announcement","url":"https://microsoft.ai/news/mai-code-1-1-flash-br-better-faster-at-a-quarter-of-the-cost/","type":"announcement"}]},"microsoft/phi-4-mini":{"id":"microsoft/phi-4-mini","name":"Phi-4-mini","description":"Compact Microsoft instruction model tuned for efficient coding assistance, reasoning, and low-latency agent tasks","family":"phi","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-02-26","last_updated":"2025-02-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"links":[{"label":"Weights","url":"https://huggingface.co/microsoft/Phi-4-mini-instruct","type":"weights"}],"benchmarks":[{"name":"MMLU","score":67.3,"metric":"accuracy","source":"https://huggingface.co/microsoft/Phi-4-mini-instruct/resolve/main/README.md"}]},"microsoft/mai-code-1-flash":{"id":"microsoft/mai-code-1-flash","name":"MAI-Code-1-Flash","description":"Microsoft coding model built for fast, efficient assistance in everyday developer workflows","family":"mai","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-12","release_date":"2026-06-02","last_updated":"2026-06-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000},"links":[{"label":"Model card","url":"https://microsoft.ai/pdf/MAI-Code-1-Flash-Model-Card.PDF","type":"model_card"},{"label":"Announcement","url":"https://microsoft.ai/news/introducingmai-code-1-flash/","type":"announcement"}],"benchmarks":[{"name":"SWE-Bench Pro","score":51.2,"metric":"resolve rate","harness":"GitHub Copilot","source":"https://microsoft.ai/news/introducingmai-code-1-flash/","date":"2026-06-02"},{"name":"SWE-Bench Verified","score":71.6,"metric":"resolved","source":"https://llm-stats.com/benchmarks/swe-bench-verified"},{"name":"Terminal-Bench","score":54.8,"metric":"success rate","version":"2.0","source":"https://llm-stats.com/benchmarks/terminal-bench-2"},{"name":"GPQA Diamond","score":84.6,"metric":"accuracy","source":"https://llm-stats.com/benchmarks/gpqa"}]},"microsoft/wizardlm-2-8x22b":{"id":"microsoft/wizardlm-2-8x22b","name":"WizardLM 2 8x22B","description":"Microsoft WizardLM instruction-tuned Mixtral MoE for multilingual dialogue and complex tasks","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2024-04-15","last_updated":"2024-04-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":8000},"license":"Apache-2.0","weights":[{"label":"Hugging Face archive","url":"https://huggingface.co/alpindale/WizardLM-2-8x22B"}]},"stepfun/step-3.7-flash":{"id":"stepfun/step-3.7-flash","name":"Step 3.7 Flash","description":"Newer StepFun flash model for faster agents, coding, and multimodal prompts","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2026-03-01","release_date":"2026-05-29","last_updated":"2026-05-29","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/stepfun-ai/Step-3.7-Flash"}],"benchmarks":[{"name":"SWE-Bench Pro","score":56.3,"metric":"resolve rate","source":"https://static.stepfun.com/blog/step-3.7-flash/","date":"2026-05-29"},{"name":"SWE-Bench Verified","score":76.5,"metric":"resolved","source":"https://static.stepfun.com/blog/step-3.7-flash/","date":"2026-05-29"},{"name":"Terminal-Bench","score":59.6,"metric":"success rate","version":"2.1","source":"https://static.stepfun.com/blog/step-3.7-flash/","date":"2026-05-29"},{"name":"Humanity's Last Exam","score":47.2,"metric":"accuracy","variant":"with tools","source":"https://static.stepfun.com/blog/step-3.7-flash/","date":"2026-05-29"},{"name":"BrowseComp","score":75.8,"metric":"accuracy","source":"https://static.stepfun.com/blog/step-3.7-flash/","date":"2026-05-29"},{"name":"Toolathlon","score":49.5,"metric":"success rate","source":"https://static.stepfun.com/blog/step-3.7-flash/","date":"2026-05-29"},{"name":"GDPval","score":45.8,"metric":"wins or ties","source":"https://static.stepfun.com/blog/step-3.7-flash/","date":"2026-05-29"},{"name":"ClawEval","score":67.1,"metric":"pass^3","version":"1.1","source":"https://static.stepfun.com/blog/step-3.7-flash/","date":"2026-05-29"},{"name":"Artificial Analysis Coding Index","score":37.1,"metric":"index","source":"https://openrouter.ai/stepfun/step-3.7-flash/benchmarks","date":"2026-06-15"},{"name":"SciCode","score":40,"metric":"percent correct","source":"https://openrouter.ai/stepfun/step-3.7-flash/benchmarks","date":"2026-06-15"},{"name":"Terminal-Bench Hard","score":35.6,"metric":"success rate","source":"https://openrouter.ai/stepfun/step-3.7-flash/benchmarks","date":"2026-06-15"}]},"stepfun/step-5-preview":{"id":"stepfun/step-5-preview","name":"Step 5 Preview","description":"StepFun's next-generation flagship base model for coding and professional knowledge work, with native text, image, and video input and a 1M-token context window","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-09-16","last_updated":"2026-09-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"input":1000000,"output":1000000},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/stepfun-ai/Step-5-Preview"}]},"stepfun/step-3.5-flash":{"id":"stepfun/step-3.5-flash","name":"Step 3.5 Flash","description":"StepFun flash lane for quick multimodal reasoning and coding assistance","family":"step","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-01-29","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/stepfun-ai/Step-3.5-Flash"}],"benchmarks":[{"name":"Artificial Analysis Coding Index","score":31.6,"metric":"index","source":"https://openrouter.ai/stepfun/step-3.5-flash/benchmarks","date":"2026-06-02"},{"name":"SciCode","score":40.4,"metric":"percent correct","source":"https://openrouter.ai/stepfun/step-3.5-flash/benchmarks","date":"2026-06-02"},{"name":"Terminal-Bench Hard","score":27.3,"metric":"success rate","source":"https://openrouter.ai/stepfun/step-3.5-flash/benchmarks","date":"2026-06-02"},{"name":"SWE-Bench Verified","score":74.4,"metric":"resolved","source":"https://arxiv.org/abs/2602.10604"}]},"stepfun/step-3.5-flash-2603":{"id":"stepfun/step-3.5-flash-2603","name":"Step 3.5 Flash 2603","description":"StepFun flash model for efficient multimodal reasoning, coding, and tool use","family":"step","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":256000,"output":256000},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/stepfun-ai/Step-3.5-Flash"}],"benchmarks":[{"name":"Artificial Analysis Coding Index","score":34.6,"metric":"index","source":"https://openrouter.ai/stepfun/step-3.5-flash/benchmarks","date":"2026-06-02"},{"name":"SciCode","score":38.5,"metric":"percent correct","source":"https://openrouter.ai/stepfun/step-3.5-flash/benchmarks","date":"2026-06-02"},{"name":"Terminal-Bench Hard","score":32.6,"metric":"success rate","source":"https://openrouter.ai/stepfun/step-3.5-flash/benchmarks","date":"2026-06-02"}]},"inception/mercury-edit-2":{"id":"inception/mercury-edit-2","name":"Mercury Edit 2","description":"Code editing dLLM for autocomplete (FIM) and next-edit suggestions","family":"mercury","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-03-30","last_updated":"2026-03-30","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32000,"output":8192}},"inception/mercury-3":{"id":"inception/mercury-3","name":"Mercury 3","description":"Inception's diffusion reasoning model for chat, multi-step problem solving, and tool use","family":"mercury","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-10-16","last_updated":"2026-10-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072}},"inception/mercury-2.5":{"id":"inception/mercury-2.5","name":"Mercury 2.5","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"mercury","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-09-08","last_updated":"2026-09-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":260000,"output":65536}},"inception/mercury-2":{"id":"inception/mercury-2","name":"Mercury 2","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"mercury","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-02-24","last_updated":"2026-02-24","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":50000}},"relace/relace-search":{"id":"relace/relace-search","name":"Relace Search","description":"Tool-capable chat model for instruction following and agentic application workflows","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-12-08","last_updated":"2025-12-08","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000}},"relace/relace-apply-3":{"id":"relace/relace-apply-3","name":"Relace Apply 3","description":"General-purpose chat model for instruction following, writing, and analysis","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-09-26","last_updated":"2025-09-26","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000}},"swiss-ai/apertus-8b":{"id":"swiss-ai/apertus-8b","name":"Apertus 8B","description":"Fully open 8B multilingual LLM supporting 1800+ languages with 65K context. Trained on compliant open data. Apache 2.0, EU AI Act compliant.","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-09-02","last_updated":"2025-09-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":8192},"license":"Apache-2.0","links":[{"label":"Paper","url":"https://arxiv.org/abs/2509.14233","type":"paper"}],"weights":[{"label":"Hugging Face","url":"https://huggingface.co/swiss-ai/Apertus-8B-Instruct-2509"}]},"swiss-ai/apertus-70b":{"id":"swiss-ai/apertus-70b","name":"Apertus 70B","description":"Fully open 70B multilingual LLM supporting 1800+ languages with 65K context. Trained on 15T tokens of compliant open data. Apache 2.0, EU AI Act compliant.","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-09-02","last_updated":"2025-09-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":8192},"license":"Apache-2.0","links":[{"label":"Paper","url":"https://arxiv.org/abs/2509.14233","type":"paper"}],"weights":[{"label":"Hugging Face","url":"https://huggingface.co/swiss-ai/Apertus-70B-Instruct-2509"}]},"deepreinforce/ornith-1.0-9b":{"id":"deepreinforce/ornith-1.0-9b","name":"Ornith 1.0 9B","description":"Open coding-reasoning model for repository tasks and self-improving agents","family":"ornith","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-06-25","last_updated":"2026-06-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144},"license":"MIT","links":[{"label":"Model card","url":"https://huggingface.co/deepreinforce-ai/Ornith-1.0-9B","type":"model_card"},{"label":"Announcement","url":"https://deep-reinforce.com/ornith_1_0.html","type":"announcement"}],"weights":[{"label":"Hugging Face","url":"https://huggingface.co/deepreinforce-ai/Ornith-1.0-9B"}],"benchmarks":[{"name":"SWE-Bench Verified","score":69.4,"metric":"percent resolved","source":"https://huggingface.co/deepreinforce-ai/Ornith-1.0-9B"},{"name":"SWE-Bench Pro","score":42.9,"metric":"percent resolved","source":"https://huggingface.co/deepreinforce-ai/Ornith-1.0-9B"},{"name":"SWE-Bench Multilingual","score":52,"metric":"percent resolved","source":"https://huggingface.co/deepreinforce-ai/Ornith-1.0-9B"},{"name":"Terminal-Bench 2.1","score":43.1,"metric":"percent","variant":"Terminus-2","source":"https://huggingface.co/deepreinforce-ai/Ornith-1.0-9B"},{"name":"Terminal-Bench 2.1","score":40.6,"metric":"percent","variant":"Claude Code","source":"https://huggingface.co/deepreinforce-ai/Ornith-1.0-9B"},{"name":"NL2Repo","score":27.2,"metric":"percent","source":"https://huggingface.co/deepreinforce-ai/Ornith-1.0-9B"},{"name":"Claw-eval","score":63.1,"metric":"percent","source":"https://huggingface.co/deepreinforce-ai/Ornith-1.0-9B"}]},"deepreinforce/ornith-1.0-31b":{"id":"deepreinforce/ornith-1.0-31b","name":"Ornith 1.0 31B","description":"Open coding-reasoning model for repository tasks and self-improving agents","family":"ornith","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-06-25","last_updated":"2026-06-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144},"license":"MIT","links":[{"label":"Announcement","url":"https://deep-reinforce.com/ornith_1_0.html","type":"announcement"}]},"deepreinforce/ornith-1.0-35b":{"id":"deepreinforce/ornith-1.0-35b","name":"Ornith 1.0 35B","description":"Large coding-reasoning model for agentic software tasks and RL search","family":"ornith","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-06-25","last_updated":"2026-06-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144},"license":"MIT","links":[{"label":"Model card","url":"https://huggingface.co/deepreinforce-ai/Ornith-1.0-35B","type":"model_card"},{"label":"Announcement","url":"https://deep-reinforce.com/ornith_1_0.html","type":"announcement"}],"weights":[{"label":"Hugging Face","url":"https://huggingface.co/deepreinforce-ai/Ornith-1.0-35B"}],"benchmarks":[{"name":"SWE-Bench Verified","score":75.6,"metric":"percent resolved","source":"https://huggingface.co/deepreinforce-ai/Ornith-1.0-35B"},{"name":"SWE-Bench Pro","score":50.4,"metric":"percent resolved","source":"https://huggingface.co/deepreinforce-ai/Ornith-1.0-35B"},{"name":"SWE-Bench Multilingual","score":69.3,"metric":"percent resolved","source":"https://huggingface.co/deepreinforce-ai/Ornith-1.0-35B"},{"name":"Terminal-Bench 2.1","score":64.2,"metric":"percent","variant":"Terminus-2","source":"https://huggingface.co/deepreinforce-ai/Ornith-1.0-35B"},{"name":"Terminal-Bench 2.1","score":62.8,"metric":"percent","variant":"Claude Code","source":"https://huggingface.co/deepreinforce-ai/Ornith-1.0-35B"},{"name":"NL2Repo","score":34.6,"metric":"percent","source":"https://huggingface.co/deepreinforce-ai/Ornith-1.0-35B"},{"name":"Claw-eval","score":69.8,"metric":"percent","source":"https://huggingface.co/deepreinforce-ai/Ornith-1.0-35B"}]},"deepreinforce/ornith-1.5-35b-a3b":{"id":"deepreinforce/ornith-1.5-35b-a3b","name":"Ornith 1.5 35B A3B","description":"Mixture-of-experts coding-reasoning model for agentic software tasks, tool use, and image understanding","family":"ornith","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-08-18","last_updated":"2026-08-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144},"license":"MIT","links":[{"label":"Model card","url":"https://huggingface.co/ornith-ai/Ornith-1.5-35B-A3B","type":"model_card"}],"weights":[{"label":"Hugging Face","url":"https://huggingface.co/ornith-ai/Ornith-1.5-35B-A3B"}]},"deepreinforce/ornith-1.0-397b":{"id":"deepreinforce/ornith-1.0-397b","name":"Ornith 1.0 397B","description":"Large coding-reasoning model for agentic software tasks and RL search","family":"ornith","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-06-25","last_updated":"2026-06-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144},"license":"MIT","links":[{"label":"Model card","url":"https://huggingface.co/deepreinforce-ai/Ornith-1.0-397B","type":"model_card"},{"label":"Announcement","url":"https://deep-reinforce.com/ornith_1_0.html","type":"announcement"}],"weights":[{"label":"Hugging Face","url":"https://huggingface.co/deepreinforce-ai/Ornith-1.0-397B"},{"label":"Hugging Face (FP8)","url":"https://huggingface.co/deepreinforce-ai/Ornith-1.0-397B-FP8","quantization":"fp8"}],"benchmarks":[{"name":"SWE-Bench Verified","score":82.4,"metric":"percent resolved","source":"https://huggingface.co/deepreinforce-ai/Ornith-1.0-397B"},{"name":"SWE-Bench Pro","score":62.2,"metric":"percent resolved","source":"https://huggingface.co/deepreinforce-ai/Ornith-1.0-397B"},{"name":"SWE-Bench Multilingual","score":78.9,"metric":"percent resolved","source":"https://huggingface.co/deepreinforce-ai/Ornith-1.0-397B"},{"name":"Terminal-Bench 2.1","score":77.5,"metric":"percent","variant":"Terminus-2","source":"https://huggingface.co/deepreinforce-ai/Ornith-1.0-397B"},{"name":"Terminal-Bench 2.1","score":78.2,"metric":"percent","variant":"Claude Code","source":"https://huggingface.co/deepreinforce-ai/Ornith-1.0-397B"},{"name":"NL2Repo","score":48.2,"metric":"percent","source":"https://huggingface.co/deepreinforce-ai/Ornith-1.0-397B"},{"name":"Claw-eval","score":77.1,"metric":"percent","source":"https://huggingface.co/deepreinforce-ai/Ornith-1.0-397B"}]},"fireworks/ember-1":{"id":"fireworks/ember-1","name":"Ember-1","description":"Specialized model from Fireworks built on Kimi K3, producing shorter reasoning traces with approximately 40% fewer tokens while maintaining comparable quality","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072}},"meituan/longcat-2.0":{"id":"meituan/longcat-2.0","name":"LongCat-2.0","description":"Meituan LongCat-2.0, a reasoning model with tool calling and a 1M-token context window","family":"longcat","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":131072},"benchmarks":[{"name":"SWE-Bench Pro","score":59.5,"metric":"resolve rate","source":"https://github.com/meituan-longcat/longcat-2.0","date":"2026-06-30"},{"name":"SWE-Bench Multilingual","score":77.3,"metric":"resolve rate","source":"https://github.com/meituan-longcat/longcat-2.0","date":"2026-06-30"},{"name":"Terminal-Bench","score":70.8,"metric":"success rate","version":"2.1","source":"https://github.com/meituan-longcat/longcat-2.0","date":"2026-06-30"},{"name":"GPQA Diamond","score":88.9,"metric":"accuracy","source":"https://github.com/meituan-longcat/longcat-2.0","date":"2026-06-30"},{"name":"BrowseComp","score":79.9,"metric":"accuracy","source":"https://github.com/meituan-longcat/longcat-2.0","date":"2026-06-30"},{"name":"IFEval","score":90,"metric":"accuracy","source":"https://github.com/meituan-longcat/longcat-2.0","date":"2026-06-30"},{"name":"FORTE","score":73.2,"metric":"success rate","source":"https://github.com/meituan-longcat/longcat-2.0","date":"2026-06-30"}]},"meituan/longcat-2.5-preview":{"id":"meituan/longcat-2.5-preview","name":"LongCat-2.5-Preview","description":"Meituan's multimodal reasoning model for coding and agent tasks, with image understanding and a 1M-token context window","family":"longcat","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"release_date":"2026-09-25","last_updated":"2026-09-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072}},"phocinae/phocinae-largha-150m-v1":{"id":"phocinae/phocinae-largha-150m-v1","type":"decision","name":"Phocinae Largha 150M v1","description":"Encoder decision model that answers typed questions (noul, choice, score) over text with calibrated probabilities, on CPU and GPU","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-08","last_updated":"2026-10-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":0},"license":"Apache-2.0","links":[{"label":"Model card","url":"https://huggingface.co/Phocinae/Phocinae-Largha-150M-v1","type":"model_card"}],"weights":[{"label":"Hugging Face","url":"https://huggingface.co/Phocinae/Phocinae-Largha-150M-v1"}]},"mistral/ministral-3b-2512":{"id":"mistral/ministral-3b-2512","name":"Ministral 3 3B 2512","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":104857}},"mistral/mistral-small-2506":{"id":"mistral/mistral-small-2506","name":"Mistral Small 3.2","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-03","release_date":"2025-06-20","last_updated":"2025-06-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/mistralai/Mistral-Small-3.2-24B-Instruct-2506"}]},"mistral/mistral-large-2512":{"id":"mistral/mistral-large-2512","name":"Mistral Large 3","description":"Mistral's largest general model for enterprise agents, coding, and multilingual reasoning","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-11","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/mistralai/Mistral-Large-3-675B-Instruct-2512"}],"benchmarks":[{"name":"Artificial Analysis Coding Index","score":22.7,"metric":"index","source":"https://openrouter.ai/mistralai/mistral-large-2512/benchmarks","date":"2026-06-04"},{"name":"SciCode","score":36.2,"metric":"percent correct","source":"https://openrouter.ai/mistralai/mistral-large-2512/benchmarks","date":"2026-06-04"},{"name":"Terminal-Bench Hard","score":15.9,"metric":"success rate","source":"https://openrouter.ai/mistralai/mistral-large-2512/benchmarks","date":"2026-06-04"}]},"mistral/magistral-medium-latest":{"id":"mistral/magistral-medium-latest","name":"Magistral Medium (latest)","description":"Mistral reasoning model for transparent analysis, math, and complex decisions","family":"magistral-medium","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2025-03-17","last_updated":"2025-03-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384}},"mistral/mistral-7b-instruct-v0.3":{"id":"mistral/mistral-7b-instruct-v0.3","name":"Mistral 7B Instruct v0.3","description":"Open 7B Mistral instruction model with function calling","family":"mistral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2024-05-22","last_updated":"2024-05-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"license":"Apache 2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.3"}]},"mistral/ministral-3-8b-instruct-2512":{"id":"mistral/ministral-3-8b-instruct-2512","name":"Ministral 3 8B","description":"Compact open vision-language model for edge deployment, instruction following, and tool use","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"license":"Apache 2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/mistralai/Ministral-3-8B-Instruct-2512"}]},"mistral/mistral-nemo":{"id":"mistral/mistral-nemo","name":"Mistral Nemo","description":"Efficient Mistral-NVIDIA open model for multilingual chat and local deployment","family":"mistral-nemo","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/mistralai/Mistral-Nemo-Instruct-2407"}]},"mistral/mistral-large-2411":{"id":"mistral/mistral-large-2411","name":"Mistral Large 2.1","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-11","release_date":"2024-11-18","last_updated":"2024-11-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/mistralai/Mistral-Large-Instruct-2411"}],"benchmarks":[{"name":"Artificial Analysis Coding Index","score":13.8,"metric":"index","source":"https://openrouter.ai/mistralai/mistral-large-2407/benchmarks","date":"2026-03-11"},{"name":"SciCode","score":29.2,"metric":"percent correct","source":"https://openrouter.ai/mistralai/mistral-large-2407/benchmarks","date":"2026-03-11"},{"name":"Terminal-Bench Hard","score":6.1,"metric":"success rate","source":"https://openrouter.ai/mistralai/mistral-large-2407/benchmarks","date":"2026-03-11"}]},"mistral/mistral-large":{"id":"mistral/mistral-large","name":"Mistral Large","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-11-30","release_date":"2024-02-26","last_updated":"2024-02-26","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":102400}},"mistral/devstral-small-2507":{"id":"mistral/devstral-small-2507","name":"Devstral Small","description":"Mistral coding agent model for repository tasks and software engineering workflows","family":"devstral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-07-10","last_updated":"2025-07-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/mistralai/Devstral-Small-2507"}],"benchmarks":[{"name":"SWE-Bench Verified","score":53.6,"metric":"resolved","source":"https://mistral.ai/news/devstral-2507","date":"2025-07-10"}]},"mistral/mistral-small-2603":{"id":"mistral/mistral-small-2603","name":"Mistral Small 4","description":"Fast Mistral production model for chat, extraction, and cost-sensitive agents","family":"mistral-small","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/mistralai/Mistral-Small-4-119B-2603"}],"benchmarks":[{"name":"Artificial Analysis Coding Index","score":24.3,"metric":"index","source":"https://openrouter.ai/mistralai/mistral-small-2603/benchmarks","date":"2026-06-01"},{"name":"SciCode","score":38,"metric":"percent correct","source":"https://openrouter.ai/mistralai/mistral-small-2603/benchmarks","date":"2026-06-01"},{"name":"Terminal-Bench Hard","score":17.4,"metric":"success rate","source":"https://openrouter.ai/mistralai/mistral-small-2603/benchmarks","date":"2026-06-01"}]},"mistral/mistral-small-3-1-24b-instruct-2503":{"id":"mistral/mistral-small-3-1-24b-instruct-2503","name":"Mistral Small 3.1 24B","description":"Efficient multimodal model for instruction following, coding, reasoning, and function calling","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-03-17","last_updated":"2025-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":16384},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/mistralai/Mistral-Small-3.1-24B-Instruct-2503"}]},"mistral/voxtral-small-latest":{"id":"mistral/voxtral-small-latest","name":"Voxtral Small (latest)","description":"Instruct model with native audio input for speech understanding and tool use","family":"voxtral","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-07-15","last_updated":"2025-07-15","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":true,"limit":{"context":32000,"output":32000}},"mistral/ministral-14b":{"id":"mistral/ministral-14b","name":"Ministral 14B","description":"Compact multimodal Mistral model for local assistants, edge agents, and efficient tool use","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"license":"Apache-2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/mistralai/Ministral-3-14B-Instruct-2512"}]},"mistral/ministral-3b":{"id":"mistral/ministral-3b","name":"Ministral 3B","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-03","release_date":"2024-10-16","last_updated":"2024-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192}},"mistral/mistral-large-4":{"id":"mistral/mistral-large-4","name":"Mistral Large 4","description":"Mistral's 1T-parameter multimodal MoE for reasoning, coding, and agentic work","family":"mistral-large","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":262144}},"mistral/codestral-latest":{"id":"mistral/codestral-latest","name":"Codestral (latest)","description":"Mistral code model for completions, refactors, and developer IDE workflows","family":"codestral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-05-29","last_updated":"2025-01-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":4096},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/mistralai/Codestral-22B-v0.1"}],"benchmarks":[{"name":"Aider Polyglot","score":11.1,"metric":"percent correct","source":"https://aider.chat/docs/leaderboards/","date":"2025-01-13"}]},"mistral/pixtral-12b":{"id":"mistral/pixtral-12b","name":"Pixtral 12B","description":"Mistral vision-language model for image understanding and multimodal chat","family":"pixtral","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-09","release_date":"2024-09-11","last_updated":"2024-09-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/mistralai/Pixtral-12B-2409"}]},"mistral/codestral-22b-v0.1":{"id":"mistral/codestral-22b-v0.1","name":"Codestral-22B-v0.1","description":"Open Mistral code model for fill-in-the-middle and 80+ programming languages","family":"codestral","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2024-05-29","last_updated":"2024-05-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":8192},"license":"Mistral AI Non-Production License","weights":[{"label":"Hugging Face","url":"https://huggingface.co/mistralai/Codestral-22B-v0.1"}]},"mistral/mistral-medium-2505":{"id":"mistral/mistral-medium-2505","name":"Mistral Medium 3","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-05-07","last_updated":"2025-05-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":131072},"benchmarks":[{"name":"Artificial Analysis Coding Index","score":13.6,"metric":"index","source":"https://openrouter.ai/mistralai/mistral-medium-3/benchmarks","date":"2026-05-30"},{"name":"SciCode","score":33.1,"metric":"percent correct","source":"https://openrouter.ai/mistralai/mistral-medium-3/benchmarks","date":"2026-05-30"},{"name":"Terminal-Bench Hard","score":3.8,"metric":"success rate","source":"https://openrouter.ai/mistralai/mistral-medium-3/benchmarks","date":"2026-05-30"}]},"mistral/mistral-small-3.2-24b-instruct":{"id":"mistral/mistral-small-3.2-24b-instruct","name":"Mistral Small 3.2 24B","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-10-31","release_date":"2025-06-20","last_updated":"2025-06-20","modalities":{"input":["image","text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":16384}},"mistral/voxtral-small-24b-2507":{"id":"mistral/voxtral-small-24b-2507","name":"Voxtral Small 24B 2507","description":"Open audio-language model for speech transcription, audio understanding, and voice-driven tool use","family":"voxtral","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-07-15","last_updated":"2025-07-15","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"license":"Apache 2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/mistralai/Voxtral-Small-24B-2507"}]},"mistral/devstral-2512":{"id":"mistral/devstral-2512","name":"Devstral 2","description":"Mistral's coding-agent model for repository work, terminal tasks, and software fixes","family":"devstral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-09","last_updated":"2025-12-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/mistralai/Devstral-2-123B-Instruct-2512"}],"benchmarks":[{"name":"Artificial Analysis Coding Index","score":23.7,"metric":"index","source":"https://openrouter.ai/mistralai/devstral-2512/benchmarks","date":"2026-05-31"},{"name":"SciCode","score":33.1,"metric":"percent correct","source":"https://openrouter.ai/mistralai/devstral-2512/benchmarks","date":"2026-05-31"},{"name":"Terminal-Bench Hard","score":18.9,"metric":"success rate","source":"https://openrouter.ai/mistralai/devstral-2512/benchmarks","date":"2026-05-31"}]},"mistral/mistral-large-2407":{"id":"mistral/mistral-large-2407","name":"Mistral Large 2407","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-03-31","release_date":"2024-07-24","last_updated":"2024-07-24","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":104857},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/mistralai/Mistral-Large-Instruct-2407"}]},"mistral/mistral-medium-latest":{"id":"mistral/mistral-medium-latest","name":"Mistral Medium (latest)","description":"Balanced Mistral model for enterprise assistants, multilingual work, and tools","family":"mistral-medium","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-04-29","last_updated":"2026-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/mistralai/Mistral-Medium-3.5-128B"}],"benchmarks":[{"name":"SWE-Bench Verified","score":77.6,"metric":"resolved","source":"https://huggingface.co/mistralai/Mistral-Medium-3.5-128B"}]},"mistral/ministral-3-14b-instruct-2512":{"id":"mistral/ministral-3-14b-instruct-2512","name":"Ministral 3 14B","description":"Open vision-language model for efficient local deployment, instruction following, and tool use","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"license":"Apache 2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/mistralai/Ministral-3-14B-Instruct-2512"}]},"mistral/ministral-14b-2512":{"id":"mistral/ministral-14b-2512","name":"Ministral 3 14B 2512","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":209715}},"mistral/mistral-embed":{"id":"mistral/mistral-embed","type":"embedding","name":"Mistral Embed","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"mistral-embed","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2023-12-11","last_updated":"2023-12-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8000,"output":3072}},"mistral/mixtral-8x22b-instruct":{"id":"mistral/mixtral-8x22b-instruct","name":"Mixtral 8x22B Instruct","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-01-31","release_date":"2024-04-17","last_updated":"2024-04-17","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":52428}},"mistral/mistral-medium-2604":{"id":"mistral/mistral-medium-2604","name":"Mistral Medium 3.5","description":"Balanced Mistral model for enterprise assistants, multilingual work, and tools","family":"mistral-medium","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-04-29","last_updated":"2026-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/mistralai/Mistral-Medium-3.5-128B"}],"benchmarks":[{"name":"SWE-Bench Verified","score":77.6,"metric":"resolved","source":"https://huggingface.co/mistralai/Mistral-Medium-3.5-128B"},{"name":"τ³-Telecom","score":91.4,"metric":"accuracy","variant":"public preview","source":"https://mistral.ai/news/vibe-remote-agents-mistral-medium-3-5/","date":"2026-05-22"}]},"mistral/magistral-small-2509":{"id":"mistral/magistral-small-2509","name":"Magistral Small 1.2","description":"Open multimodal reasoning model for transparent analysis of text and images","family":"magistral","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2025-09-18","last_updated":"2025-09-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"license":"Apache 2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/mistralai/Magistral-Small-2509"}]},"mistral/pixtral-large-2502":{"id":"mistral/pixtral-large-2502","name":"Pixtral Large (25.02)","description":"Mistral vision-language model for image understanding and multimodal chat","family":"pixtral","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-04-08","last_updated":"2025-04-08","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192}},"mistral/devstral-medium-2507":{"id":"mistral/devstral-medium-2507","name":"Devstral Medium","description":"Mistral coding agent model for repository tasks and software engineering workflows","family":"devstral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-07-10","last_updated":"2025-07-10","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":128000},"benchmarks":[{"name":"SWE-Bench Verified","score":61.6,"metric":"resolved","source":"https://mistral.ai/news/devstral-2507","date":"2025-07-10"}]},"mistral/mistral-medium-3.1":{"id":"mistral/mistral-medium-3.1","name":"Mistral Medium 3.1","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-06-30","release_date":"2025-08-12","last_updated":"2025-08-12","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":104857}},"mistral/voxtral-mini-3b-2507":{"id":"mistral/voxtral-mini-3b-2507","name":"Voxtral Mini 3B 2507","description":"Open audio-language model for speech transcription, audio understanding, and voice-driven tool use","family":"voxtral","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-07-15","last_updated":"2025-07-15","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"license":"Apache 2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/mistralai/Voxtral-Mini-3B-2507"}]},"mistral/mistral-saba":{"id":"mistral/mistral-saba","name":"Saba","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-02-17","last_updated":"2025-02-17","modalities":{"input":["text","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":26214}},"mistral/devstral-small-2":{"id":"mistral/devstral-small-2","name":"Devstral Small 2","description":"Compact multimodal coding model for repository exploration, file editing, and software agents","family":"devstral","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-09","last_updated":"2025-12-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"license":"Apache-2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/mistralai/Devstral-Small-2-24B-Instruct-2512"}]},"mistral/ministral-3-3b-instruct-2512":{"id":"mistral/ministral-3-3b-instruct-2512","name":"Ministral 3 3B","description":"Compact open vision-language model for edge deployment, instruction following, and tool use","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"license":"Apache 2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/mistralai/Ministral-3-3B-Instruct-2512"}]},"mistral/mistral-medium-3":{"id":"mistral/mistral-medium-3","name":"Mistral Medium 3","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-07","last_updated":"2025-05-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":104857}},"mistral/ministral-8b-instruct-2410":{"id":"mistral/ministral-8b-instruct-2410","name":"Ministral 8B Instruct","description":"Efficient open Mistral edge model for on-device chat and function calling","family":"ministral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2024-10-16","last_updated":"2024-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"license":"Mistral Research License","weights":[{"label":"Hugging Face","url":"https://huggingface.co/mistralai/Ministral-8B-Instruct-2410"}]},"mistral/ministral-8b-2512":{"id":"mistral/ministral-8b-2512","name":"Ministral 3 8B 2512","description":"Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads","family":"ministral","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":209715}},"mistral/mistral-large-latest":{"id":"mistral/mistral-large-latest","name":"Mistral Large (latest)","description":"Flagship Mistral model for advanced reasoning, coding, and multilingual work","family":"mistral-large","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-11","release_date":"2024-11-01","last_updated":"2025-12-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/mistralai/Mistral-Large-3-675B-Instruct-2512"}]},"mistral/mistral-small-24b-instruct-2501":{"id":"mistral/mistral-small-24b-instruct-2501","name":"Mistral Small 3","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2023-10-31","release_date":"2025-01-30","last_updated":"2025-01-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":16384}},"mistral/mistral-small-3.1-24b-instruct":{"id":"mistral/mistral-small-3.1-24b-instruct","name":"Mistral Small 3.1 24B","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2023-10-31","release_date":"2025-03-17","last_updated":"2025-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":102400}},"mistral/codestral-2508":{"id":"mistral/codestral-2508","name":"Codestral 25.08","description":"Mistral code model for low-latency completion, fill-in-the-middle, and code generation","family":"codestral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-07-30","last_updated":"2025-07-30","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192}},"mistral/magistral-small-2506":{"id":"mistral/magistral-small-2506","name":"Magistral Small","description":"Open Mistral reasoning model for transparent step-by-step problem solving","family":"magistral","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2025-06-10","last_updated":"2025-06-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"license":"Apache 2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/mistralai/Magistral-Small-2506"}]},"mistral/pixtral-large-latest":{"id":"mistral/pixtral-large-latest","name":"Pixtral Large (latest)","description":"Mistral's larger vision model for document-heavy image understanding and chat","family":"pixtral","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-11","release_date":"2024-11-01","last_updated":"2024-11-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/mistralai/Pixtral-Large-Instruct-2411"}]},"mistral/mistral-small-latest":{"id":"mistral/mistral-small-latest","name":"Mistral Small (latest)","description":"Efficient Mistral model for fast chat, extraction, and production assistants","family":"mistral-small","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":256000},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/mistralai/Mistral-Small-4-119B-2603"}]},"mistral/devstral-medium-latest":{"id":"mistral/devstral-medium-latest","name":"Devstral 2 (latest)","description":"Mistral coding agent model for repository tasks and software engineering workflows","family":"devstral","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-12","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/mistralai/Devstral-2-123B-Instruct-2512"}]},"mistral/mixtral-8x7b-instruct-v0.1":{"id":"mistral/mixtral-8x7b-instruct-v0.1","name":"Mixtral 8x7B Instruct v0.1","description":"Open sparse mixture-of-experts Mistral instruction model","family":"mistral","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2023-12-11","last_updated":"2023-12-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"license":"Apache 2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/mistralai/Mixtral-8x7B-Instruct-v0.1"}]},"mistral/mistral-medium-3-5":{"id":"mistral/mistral-medium-3-5","name":"Mistral Medium 3.5","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral-medium","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-04-30","last_updated":"2026-04-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":209715}},"anthracite-org/magnum-v4-72b":{"id":"anthracite-org/magnum-v4-72b","name":"Magnum v4 72B","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-06-30","release_date":"2024-10-22","last_updated":"2024-10-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":4096}},"black-forest-labs/flux-1-dev":{"id":"black-forest-labs/flux-1-dev","type":"image","name":"FLUX.1 [dev]","description":"Open-weight FLUX.1 text-to-image model for high-quality guidance-distilled generation","family":"flux","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-08-01","last_updated":"2024-08-01","modalities":{"input":["text"],"output":["image"]},"open_weights":true,"limit":{"context":0,"output":0},"license":"FLUX.1 [dev] Non-Commercial License","weights":[{"label":"Hugging Face","url":"https://huggingface.co/black-forest-labs/FLUX.1-dev"}]},"black-forest-labs/flux-1-schnell":{"id":"black-forest-labs/flux-1-schnell","type":"image","name":"FLUX.1 [schnell]","description":"Fast few-step FLUX.1 text-to-image model under Apache 2.0","family":"flux","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2024-08-01","last_updated":"2024-08-01","modalities":{"input":["text"],"output":["image"]},"open_weights":true,"limit":{"context":0,"output":0},"license":"Apache 2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/black-forest-labs/FLUX.1-schnell"}]},"black-forest-labs/flux-1-kontext-dev":{"id":"black-forest-labs/flux-1-kontext-dev","type":"image","name":"FLUX.1 Kontext [dev]","description":"Open-weight FLUX.1 Kontext model for instruction-based image editing","family":"flux","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-05-29","last_updated":"2025-05-29","modalities":{"input":["text","image"],"output":["image"]},"open_weights":true,"limit":{"context":0,"output":0},"license":"FLUX.1 [dev] Non-Commercial License","weights":[{"label":"Hugging Face","url":"https://huggingface.co/black-forest-labs/FLUX.1-Kontext-dev"}]},"bytedance-seed/seed-1-6-vision":{"id":"bytedance-seed/seed-1-6-vision","name":"Seed 1.6 Vision","description":"ByteDance Seed multimodal model for image understanding, visual reasoning, and tool-assisted tasks","family":"seed","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-08-15","last_updated":"2025-08-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000}},"bytedance-seed/seed-2.1-turbo":{"id":"bytedance-seed/seed-2.1-turbo","name":"Seed 2.1 Turbo","description":"Faster ByteDance Seed 2.1 model for multimodal reasoning and latency-sensitive agent workflows","family":"seed","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000}},"bytedance-seed/seed-2.0-lite":{"id":"bytedance-seed/seed-2.0-lite","name":"Seed 2.0 Lite","description":"Cost-efficient ByteDance Seed 2.0 model for production chat, analysis, and structured generation","family":"seed","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000}},"bytedance-seed/seedance-2.5":{"id":"bytedance-seed/seedance-2.5","type":"video","name":"Seedance 2.5","description":"ByteDance multimodal audio-video model with up to 30-second single-pass generation, multi-round extension, reference inputs and timestamp-level editing.","family":"seed","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-07-31","last_updated":"2026-10-04","modalities":{"input":["text","image","video","audio"],"output":["video","audio"]},"open_weights":false,"limit":{"context":0,"output":0},"links":[{"url":"https://seed.bytedance.com/en/blog/one-take-creation-flexible-referencing-introducing-seedance-2-5","type":"announcement"}]},"bytedance-seed/seed-evolving":{"id":"bytedance-seed/seed-evolving","name":"Seed Evolving","description":"Rolling ByteDance Seed model for rapidly updated reasoning, coding, and agent capabilities","family":"seed","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000}},"bytedance-seed/seed-2.1-pro":{"id":"bytedance-seed/seed-2.1-pro","name":"Seed 2.1 Pro","description":"Flagship ByteDance Seed 2.1 model for complex multimodal reasoning, coding, and agents","family":"seed","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000}},"bytedance-seed/seed-character":{"id":"bytedance-seed/seed-character","name":"Seed Character","description":"ByteDance Seed model optimized for character-driven dialogue and consistent conversational behavior","family":"seed","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-06-23","last_updated":"2026-06-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":256000}},"bytedance-seed/seed-2.0-code":{"id":"bytedance-seed/seed-2.0-code","name":"Seed 2.0 Code","description":"ByteDance Seed coding model for multimodal software engineering and long-running agents","family":"seed","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072}},"bytedance-seed/seed-1-6":{"id":"bytedance-seed/seed-1-6","name":"Seed 1.6","description":"ByteDance Seed model for long-context reasoning, instruction following, and tool-assisted tasks","family":"seed","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":64000}},"bytedance-seed/seed-2.0-pro":{"id":"bytedance-seed/seed-2.0-pro","name":"Seed 2.0 Pro","description":"Flagship ByteDance Seed 2.0 model for complex multimodal reasoning and long-horizon agent workflows","family":"seed","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":128000}},"bytedance-seed/seed-oss-36b-instruct":{"id":"bytedance-seed/seed-oss-36b-instruct","name":"Seed-OSS-36B-Instruct","description":"Open ByteDance Seed model for long-context reasoning, tool use, and agents","family":"seed","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-08-20","last_updated":"2025-08-20","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":524288,"output":524288},"license":"Apache 2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/ByteDance-Seed/Seed-OSS-36B-Instruct"}]},"bytedance-seed/seed-1-8":{"id":"bytedance-seed/seed-1-8","name":"Seed 1.8","description":"ByteDance Seed model for multimodal reasoning, long-context analysis, and agent workflows","family":"seed","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2025-12-28","last_updated":"2025-12-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":64000}},"bytedance-seed/seedance-2.0":{"id":"bytedance-seed/seedance-2.0","type":"video","name":"Seedance 2.0","description":"ByteDance unified multimodal audio-video model supporting text, image, video and audio references, editing, extension and synchronized stereo sound.","family":"seed","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-02-12","last_updated":"2026-10-04","modalities":{"input":["text","image","video","audio"],"output":["video","audio"]},"open_weights":false,"limit":{"context":0,"output":0},"links":[{"url":"https://seed.bytedance.com/en/blog/seedance-2-0-official-launch","type":"announcement"}]},"bytedance-seed/seed-1-6-flash":{"id":"bytedance-seed/seed-1-6-flash","name":"Seed 1.6 Flash","description":"Low-latency ByteDance Seed model for high-throughput chat, extraction, and lightweight tool use","family":"seed","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-08-28","last_updated":"2025-08-28","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000}},"bytedance-seed/seed-2.0-mini":{"id":"bytedance-seed/seed-2.0-mini","name":"Seed 2.0 Mini","description":"Lightweight ByteDance Seed 2.0 model for low-latency multimodal reasoning and high-volume tasks","family":"seed","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-02-14","last_updated":"2026-02-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":256000,"output":32000}},"mixedbread/toast-1":{"id":"mixedbread/toast-1","name":"Toast 1","description":"Specialized search model for knowledge-intensive questions, multi-step retrieval, and evidence synthesis","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-13","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131000,"output":4000}},"thedrummer/unslopnemo-12b":{"id":"thedrummer/unslopnemo-12b","name":"UnslopNemo 12B","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-04-30","release_date":"2024-11-08","last_updated":"2024-11-08","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":819200}},"thedrummer/cydonia-24b-v4.1":{"id":"thedrummer/cydonia-24b-v4.1","name":"Cydonia 24B V4.1","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-04-30","release_date":"2025-09-27","last_updated":"2025-09-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964}},"thedrummer/skyfall-36b-v2":{"id":"thedrummer/skyfall-36b-v2","name":"Skyfall 36B V2","description":"Open-weight instruction model for adaptable chat and self-hosted production workloads","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-06-30","release_date":"2025-03-10","last_updated":"2025-03-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":29491}},"nace-ai/drex-v1.5":{"id":"nace-ai/drex-v1.5","type":"decision","name":"Drex 1.5","description":"System One decision model for typed, calibrated probabilistic answers over text or structured state","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":139264,"output":0},"license":"Nace.AI Open RAIL-M","links":[{"label":"Documentation","url":"https://drex.nace.ai/docs","type":"docs"},{"label":"License","url":"https://huggingface.co/nace-ai/drex-v1.5/blob/main/LICENSE","type":"license"}],"weights":[{"url":"https://huggingface.co/nace-ai/drex-v1.5","format":"safetensors"}]},"nace-ai/drex-v2.0":{"id":"nace-ai/drex-v2.0","type":"decision","name":"Drex 2.0","description":"35B mixture-of-experts System One decision model with 3B active parameters, returning typed, calibrated probabilities over text or structured state without generating text","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-06","last_updated":"2026-10-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":0},"license":"Nace.AI Open RAIL-M","links":[{"label":"Documentation","url":"https://github.com/nace-ai/drex-decision-models","type":"docs"},{"label":"License","url":"https://huggingface.co/nace-ai/drex-v2.0/blob/main/LICENSE","type":"license"}],"weights":[{"url":"https://huggingface.co/nace-ai/drex-v2.0","format":"safetensors"}]},"ibm/granite-4-h-small":{"id":"ibm/granite-4-h-small","name":"Granite-4.0-H-Small","description":"Open-weight hybrid model for enterprise chat, coding, retrieval-augmented generation, and tool-calling workloads","family":"granite","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-10-02","last_updated":"2025-10-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/ibm-granite/granite-4.0-h-small"}]},"ibm/granite-4-h-micro":{"id":"ibm/granite-4-h-micro","name":"Granite-4.0-H-Micro","description":"Compact open-weight hybrid Granite model for lightweight enterprise chat and tool calling","family":"granite","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-10-02","last_updated":"2025-10-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/ibm-granite/granite-4.0-h-micro"}]},"ibm/granite-4.2-8b":{"id":"ibm/granite-4.2-8b","name":"Granite 4.2 8B","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"granite","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-08-25","last_updated":"2026-08-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964}},"ai21/jamba-large":{"id":"ai21/jamba-large","name":"Jamba Large","description":"AI21's hybrid SSM-Transformer long-context model for enterprise agents and grounded generation","family":"jamba","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-08-22","release_date":"2025-07-03","last_updated":"2025-07-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":4096},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/ai21labs/AI21-Jamba-Large-1.7"}]},"ai21/jamba-mini":{"id":"ai21/jamba-mini","name":"Jamba Mini","description":"AI21's efficient, lightweight hybrid SSM-Transformer model for a wide range of tasks","family":"jamba","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-08-22","release_date":"2026-01-08","last_updated":"2026-01-08","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":4096},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/ai21labs/AI21-Jamba2-Mini"}]},"lightonai/lightonocr-3-0.8b":{"id":"lightonai/lightonocr-3-0.8b","name":"LightOnOCR-3 0.8B","description":"LightOnOCR-3-0.8B is the lightest model of LightOn's LightOnOCR-3 family, with the same layout and visual understanding features as its larger siblings.","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-10-08","last_updated":"2026-10-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"license":"apache-2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/lightonai/LightOnOCR-3-0.8B"}],"benchmarks":[{"name":"olmOCR-bench","score":85.5,"metric":"score","source":"https://huggingface.co/lightonai/LightOnOCR-3-0.8B"},{"name":"ParseBench","score":74.6,"metric":"mean","source":"https://huggingface.co/lightonai/LightOnOCR-3-0.8B"}]},"baidu/ernie-4.5-vl-424b-a47b":{"id":"baidu/ernie-4.5-vl-424b-a47b","name":"ERNIE 4.5 VL 424B A47B ","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"ernie","attachment":true,"reasoning":true,"tool_call":false,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-06-30","last_updated":"2025-06-30","modalities":{"input":["image","text"],"output":["text"]},"open_weights":true,"limit":{"context":123000,"output":16000}},"baidu/ernie-4.5-21b-a3b":{"id":"baidu/ernie-4.5-21b-a3b","name":"ERNIE 4.5 21B A3B","description":"Open ERNIE text MoE with 21B total and 3B active parameters","family":"ernie","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-06-30","last_updated":"2025-06-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8000},"license":"Apache-2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/baidu/ERNIE-4.5-21B-A3B-PT"}]},"baidu/cobuddy":{"id":"baidu/cobuddy","name":"CoBuddy","description":"Baidu Qianfan model optimized for code generation and tool-using coding agents","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-05-06","last_updated":"2026-05-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":65536}},"baidu/ernie-4.5-300b-a47b":{"id":"baidu/ernie-4.5-300b-a47b","name":"ERNIE 4.5 300B A47B","description":"Open ERNIE text MoE with 300B total and 47B active parameters","family":"ernie","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-06-30","last_updated":"2025-06-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":12000},"license":"Apache-2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/baidu/ERNIE-4.5-300B-A47B-Paddle"}]},"liquid/lfm2-24b-a2b":{"id":"liquid/lfm2-24b-a2b","name":"LFM2 24B A2B","description":"Efficient open Liquid MoE instruct model for on-device chat and tool calling","family":"liquid","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-02-24","last_updated":"2026-02-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"license":"LFM Open License v1.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/LiquidAI/LFM2-24B-A2B"}]},"liquid/d1":{"id":"liquid/d1","type":"decision","name":"Liquid d1","description":"System One decision model for typed classification, routing, and scoring over shared state","family":"liquid","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":65536,"output":0}},"nex-agi/nex-n2-pro":{"id":"nex-agi/nex-n2-pro","name":"Nex-N2-Pro","description":"Open agentic MoE model (397B total, 17B active) for coding, tool use, and research workflows","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/nex-agi/Nex-N2-Pro"}]},"tencent/hy-mt2-7b":{"id":"tencent/hy-mt2-7b","name":"Hy-MT2-7B","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":4096}},"tencent/hy3":{"id":"tencent/hy3","name":"Hy3","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"input":192000,"output":128000},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/tencent/Hy3"}],"benchmarks":[{"name":"SWE-Bench Verified","score":78,"metric":"resolved","source":"https://huggingface.co/tencent/Hy3"},{"name":"SWE-Bench Multilingual","score":75.8,"metric":"score","harness":"SWE-agent","variant":"highest reasoning effort","source":"https://huggingface.co/tencent/Hy3"},{"name":"SWE-Bench Pro","score":57.9,"metric":"score","harness":"SWE-agent","variant":"highest reasoning effort","source":"https://huggingface.co/tencent/Hy3"},{"name":"Terminal-Bench","score":71.7,"metric":"score","harness":"Terminus 2","variant":"highest reasoning effort; 4h timeout; 500 episodes","version":"2.1","source":"https://huggingface.co/tencent/Hy3"},{"name":"NL2Repo","score":45.6,"metric":"score","harness":"Claude Code","variant":"highest reasoning effort; 250 turns; 12000s timeout","source":"https://huggingface.co/tencent/Hy3"},{"name":"DeepSWE","score":28,"metric":"score","harness":"mini-swe-agent","variant":"highest reasoning effort; 2h timeout","source":"https://huggingface.co/tencent/Hy3"},{"name":"BrowseComp","score":84.2,"metric":"score","harness":"Tencent internal search harness","variant":"highest reasoning effort","source":"https://huggingface.co/tencent/Hy3"},{"name":"WideSearch","score":76.4,"metric":"score","harness":"Tencent internal search harness","variant":"highest reasoning effort","source":"https://huggingface.co/tencent/Hy3"},{"name":"DeepSearchQA","score":91,"metric":"score","harness":"Tencent internal search harness","variant":"highest reasoning effort","source":"https://huggingface.co/tencent/Hy3"},{"name":"MCP Atlas","score":79.1,"metric":"score","harness":"Scale AI","variant":"highest reasoning effort; April 2026; 100 tool calls","dataset":"500 public tasks","source":"https://huggingface.co/tencent/Hy3"},{"name":"Toolathlon","score":48.5,"metric":"score","variant":"highest reasoning effort","source":"https://huggingface.co/tencent/Hy3"},{"name":"APEX-Agents","score":25.6,"metric":"pass@1","variant":"highest reasoning effort","source":"https://huggingface.co/tencent/Hy3"},{"name":"ClawEval","score":68.5,"metric":"pass@3","harness":"Tencent internal harness","variant":"highest reasoning effort","dataset":"105 queries","version":"20260325","source":"https://huggingface.co/tencent/Hy3"},{"name":"WildClawBench","score":53.6,"metric":"score","harness":"OpenClaw","variant":"highest reasoning effort","dataset":"35 text-only tasks","source":"https://huggingface.co/tencent/Hy3"},{"name":"SkillsBench","score":55.3,"metric":"average over 3 runs","harness":"Claude Code","variant":"highest reasoning effort","dataset":"79 text-only tasks","source":"https://huggingface.co/tencent/Hy3"},{"name":"Humanity's Last Exam","score":53.2,"metric":"score","variant":"highest reasoning effort; with tools","dataset":"text-only","source":"https://huggingface.co/tencent/Hy3"},{"name":"Humanity's Last Exam","score":37,"metric":"score","variant":"highest reasoning effort; without tools","dataset":"text-only","source":"https://huggingface.co/tencent/Hy3"},{"name":"GPQA Diamond","score":90.4,"metric":"score","variant":"highest reasoning effort","source":"https://huggingface.co/tencent/Hy3"},{"name":"FrontierScience","score":21.3,"metric":"score","variant":"highest reasoning effort; research","source":"https://huggingface.co/tencent/Hy3"},{"name":"FrontierScience","score":74.8,"metric":"score","variant":"highest reasoning effort; olympiad","source":"https://huggingface.co/tencent/Hy3"},{"name":"USAMO","score":72,"metric":"score","variant":"highest reasoning effort","version":"2026","source":"https://huggingface.co/tencent/Hy3"},{"name":"MathArena Apex","score":38.7,"metric":"score","variant":"highest reasoning effort","source":"https://huggingface.co/tencent/Hy3"},{"name":"ArxivMath","score":52.2,"metric":"score","variant":"highest reasoning effort","source":"https://huggingface.co/tencent/Hy3"},{"name":"HorizonMath","score":7.1,"metric":"pass@12","variant":"highest reasoning effort","source":"https://huggingface.co/tencent/Hy3"},{"name":"PHYBench","score":77.4,"metric":"score","variant":"highest reasoning effort","source":"https://huggingface.co/tencent/Hy3"},{"name":"CMT Benchmark","score":37.8,"metric":"score","variant":"highest reasoning effort","source":"https://huggingface.co/tencent/Hy3"},{"name":"IMOAnswerBench","score":90,"metric":"score","variant":"highest reasoning effort","source":"https://huggingface.co/tencent/Hy3"},{"name":"SuperChem","score":54.9,"metric":"score","variant":"highest reasoning effort","source":"https://huggingface.co/tencent/Hy3"},{"name":"CL-bench","score":23.8,"metric":"score","variant":"highest reasoning effort","source":"https://huggingface.co/tencent/Hy3"},{"name":"CL-bench-life","score":17,"metric":"score","variant":"highest reasoning effort","source":"https://huggingface.co/tencent/Hy3"},{"name":"AA-LCR","score":73.4,"metric":"score","variant":"highest reasoning effort","source":"https://huggingface.co/tencent/Hy3"}]},"tencent/hy-mt2-1.8b":{"id":"tencent/hy-mt2-1.8b","name":"Hy-MT2-1.8B","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":4096}},"tencent/hy4-preview":{"id":"tencent/hy4-preview","name":"Hy4 preview","description":"A next-generation productivity model with significantly enhanced Agent and complex task execution capabilities.","family":"Hy","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-08-28","last_updated":"2026-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1024000,"output":64000}},"tencent/hy-mt2-30b-a3b":{"id":"tencent/hy-mt2-30b-a3b","name":"Hy-MT2-30B-A3B","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":4096}},"tencent/hy3-preview":{"id":"tencent/hy3-preview","name":"Hy3 preview","description":"Tencent Hy reasoning model for coding, instruction following, and agent tasks","family":"Hy","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/tencent/Hy3-preview"}],"benchmarks":[{"name":"SWE-Bench Verified","score":74.4,"metric":"resolved","source":"https://huggingface.co/tencent/Hy3-preview"}]},"aisingapore/gemma-sea-lion-v4-27b-it":{"id":"aisingapore/gemma-sea-lion-v4-27b-it","name":"Gemma-SEA-LION-v4-27B-IT","description":"Gemma 3 27B tuned by AI Singapore for Southeast Asian languages and instruction following","family":"gemma","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/aisingapore/Gemma-SEA-LION-v4-27B-IT"}]},"amazon/nova-pro-v1":{"id":"amazon/nova-pro-v1","name":"Nova Pro 1.0","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"nova-pro","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10-31","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":5120}},"amazon/nova-micro-v1":{"id":"amazon/nova-micro-v1","name":"Nova Micro 1.0","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-micro","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10-31","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":5120}},"amazon/nova-2-lite":{"id":"amazon/nova-2-lite","name":"Nova 2 Lite","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"nova","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-10","release_date":"2025-12-02","last_updated":"2025-12-01","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536}},"amazon/nova-lite-v1":{"id":"amazon/nova-lite-v1","name":"Nova Lite 1.0","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-lite","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10-31","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":5120}},"amazon/nova-pro":{"id":"amazon/nova-pro","name":"Nova Pro","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"nova-pro","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":10000}},"amazon/nova-2-lite-v1":{"id":"amazon/nova-2-lite-v1","name":"Nova 2 Lite","description":"Multimodal reasoning model for visual analysis, planning, and tool use","family":"nova","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-10","release_date":"2025-12-02","last_updated":"2025-12-02","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65535}},"amazon/nova-micro":{"id":"amazon/nova-micro","name":"Nova Micro","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-micro","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":10000}},"amazon/nova-premier":{"id":"amazon/nova-premier","name":"Nova Premier","description":"Multimodal model for complex analysis, long-context understanding, tool use, and model distillation","family":"nova","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2025-04-30","last_updated":"2025-04-30","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":10000}},"amazon/nova-lite":{"id":"amazon/nova-lite","name":"Nova Lite","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"nova-lite","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-10","release_date":"2024-12-03","last_updated":"2024-12-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":300000,"output":10000}},"apodex/apodex-1.1-mini":{"id":"apodex/apodex-1.1-mini","name":"Apodex 1.1 Mini","description":"Open-weight 35B MoE model for long-horizon research, forecasting, and tool-assisted work","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-08-24","last_updated":"2026-08-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"license":"Apache-2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/apodex/Apodex-1.1-mini"}]},"nousresearch/hermes-3-llama-3.1-405b":{"id":"nousresearch/hermes-3-llama-3.1-405b","name":"Hermes 3 405B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"nousresearch","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-08-16","last_updated":"2024-08-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384}},"nousresearch/hermes-2-pro-llama-3-8b":{"id":"nousresearch/hermes-2-pro-llama-3-8b","name":"Hermes 2 Pro Llama 3 8B","description":"Llama 3 fine-tune optimized for function calling, structured JSON responses, and dialogue","family":"hermes","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2024-04-30","last_updated":"2024-04-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":8192},"license":"Llama 3 Community License","weights":[{"label":"Hugging Face","url":"https://huggingface.co/NousResearch/Hermes-2-Pro-Llama-3-8B"}]},"nousresearch/hermes-3-llama-3.1-70b":{"id":"nousresearch/hermes-3-llama-3.1-70b","name":"Hermes 3 70B Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"nousresearch","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-08-18","last_updated":"2024-08-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384}},"nousresearch/hermes-4-405b":{"id":"nousresearch/hermes-4-405b","name":"Hermes 4 405B","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"hermes","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2024-08-31","release_date":"2025-08-26","last_updated":"2025-08-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":117964}},"inference-net/schematron-v2-small":{"id":"inference-net/schematron-v2-small","name":"Schematron V2 Small","description":"Efficient model for low-latency assistance, extraction, and routine automation","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096}},"inference-net/schematron-v2-turbo":{"id":"inference-net/schematron-v2-turbo","name":"Schematron V2 Turbo","description":"Efficient model for low-latency assistance, extraction, and routine automation","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192}},"sao10k/l3-lunaris-8b":{"id":"sao10k/l3-lunaris-8b","name":"Llama 3 8B Lunaris","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-08-13","last_updated":"2024-08-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":7372}},"sao10k/l3.3-euryale-70b":{"id":"sao10k/l3.3-euryale-70b","name":"Llama 3.3 Euryale 70B","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-12-18","last_updated":"2024-12-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384}},"sao10k/l3.1-euryale-70b":{"id":"sao10k/l3.1-euryale-70b","name":"Llama 3.1 Euryale 70B v2.2","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12-31","release_date":"2024-08-28","last_updated":"2024-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384}},"perceptron/perceptron-mk1.5":{"id":"perceptron/perceptron-mk1.5","name":"Perceptron Mk1.5","description":"Image, video, and audio understanding with visual grounding, reasoning, and function calling","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-09-25","last_updated":"2026-09-25","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":36864,"output":8192}},"perceptron/perceptron-mk1":{"id":"perceptron/perceptron-mk1","name":"Perceptron Mk1","description":"Multimodal reasoning model for visual analysis, planning, and tool use","attachment":true,"reasoning":true,"tool_call":false,"temperature":true,"release_date":"2026-05-12","last_updated":"2026-05-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":8192}},"baichuan/baichuan-m2-32b":{"id":"baichuan/baichuan-m2-32b","name":"Baichuan M2 32B","description":"Open medical model trained for clinical reasoning and patient interaction with on, off, and auto thinking modes","family":"baichuan","attachment":false,"reasoning":true,"tool_call":false,"temperature":true,"release_date":"2025-08-13","last_updated":"2025-08-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"license":"Apache-2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/baichuan-inc/Baichuan-M2-32B"}]},"upstage/solar-pro4":{"id":"upstage/solar-pro4","name":"Solar Pro 4","description":"Upstage's flagship model, specialized for agentic use","family":"solar-pro","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2026-02","release_date":"2026-08-06","last_updated":"2026-08-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":131072}},"upstage/solar-10.7b-instruct-v1.0":{"id":"upstage/solar-10.7b-instruct-v1.0","name":"SOLAR 10.7B Instruct v1.0","description":"Upstage depth-upscaled 10.7B instruction model for single-turn chat","family":"solar","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2023-12-12","last_updated":"2023-12-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":4096,"output":4096},"license":"CC-BY-NC-4.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/upstage/SOLAR-10.7B-Instruct-v1.0"}]},"upstage/solar-mini4":{"id":"upstage/solar-mini4","name":"Solar Mini 4","description":"Upstage's compact, cost-efficient model for agentic use","family":"solar-mini","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2026-02","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":524288,"output":131072}},"upstage/solar-pro2":{"id":"upstage/solar-pro2","name":"Solar Pro 2","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"solar-pro","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-03","release_date":"2025-05-20","last_updated":"2025-05-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":65536,"output":8192}},"upstage/solar-pro3":{"id":"upstage/solar-pro3","name":"Solar Pro 3","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"solar-pro","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-07","release_date":"2026-01-26","last_updated":"2026-03-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":8192}},"levanto/levanto-sage-v1.4":{"id":"levanto/levanto-sage-v1.4","type":"decision","name":"Sage 1.4","description":"Decision model that answers typed yes/no, choice, scale, sort and tag questions with calibrated probabilities","attachment":true,"reasoning":true,"tool_call":false,"temperature":false,"release_date":"2026-10-04","last_updated":"2026-10-04","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":0},"links":[{"label":"Documentation","url":"https://docs.levanto.ai/","type":"docs"}]},"thinkingmachines/inkling-small":{"id":"thinkingmachines/inkling-small","name":"Inkling Small","description":"Multimodal MoE reasoning model (276B total, 12B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-07-30","last_updated":"2026-07-30","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"license":"Apache-2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/thinkingmachines/Inkling-Small"}]},"thinkingmachines/inkling":{"id":"thinkingmachines/inkling","name":"Inkling","description":"Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio","family":"ling","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":1048576},"license":"Apache-2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/thinkingmachines/Inkling"}]},"openbmb/minicpm-v-4.7-35b-a3b":{"id":"openbmb/minicpm-v-4.7-35b-a3b","name":"MiniCPM-V 4.7 35B-A3B","description":"First MoE MiniCPM-V model for single-image, multi-image and video understanding","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"license":"Apache-2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/openbmb/MiniCPM-V-4.7-35B-A3B"}]},"openbmb/minicpm5-2b":{"id":"openbmb/minicpm5-2b","name":"MiniCPM5-2B","description":"Dense 2B-class open-source model for on-device and resource-constrained use, with native long-context support, tool calling, and agentic tasks","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-09-06","last_updated":"2026-09-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"license":"apache-2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/openbmb/MiniCPM5-2B"}]},"openbmb/minicpm5-1b":{"id":"openbmb/minicpm5-1b","name":"MiniCPM5-1B","description":"Dense 1B-class open-source model for on-device and resource-constrained use, with native long-context support, Think / No Think chat modes, and tool calling","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-05-19","last_updated":"2026-05-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"license":"apache-2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/openbmb/MiniCPM5-1B"}]},"openbmb/minicpm-v-4.7-1b":{"id":"openbmb/minicpm-v-4.7-1b","name":"MiniCPM-V 4.7 1B","description":"Compact MiniCPM-V model for single-image, multi-image and video understanding","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-10-06","last_updated":"2026-10-06","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"license":"Apache-2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/openbmb/MiniCPM-V-4.7-1B"}]},"abacusai/dracarys-llama-3.1-70b-instruct":{"id":"abacusai/dracarys-llama-3.1-70b-instruct","name":"Dracarys Llama 3.1 70B Instruct","description":"Abacus.AI coding fine-tune of Llama 3.1 70B Instruct","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2024-08-14","last_updated":"2024-08-14","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"license":"Llama 3 Community License","weights":[{"label":"Hugging Face","url":"https://huggingface.co/abacusai/Dracarys-Llama-3.1-70B-Instruct"}]},"meta/llama-3.3-70b-instruct":{"id":"meta/llama-3.3-70b-instruct","name":"Llama-3.3-70B-Instruct","description":"Popular open Llama workhorse for multilingual chat, coding, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-12-06","last_updated":"2024-12-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/meta-llama/Llama-3.3-70B-Instruct"}],"benchmarks":[{"name":"Artificial Analysis Coding Index","score":10.7,"metric":"index","source":"https://openrouter.ai/meta-llama/llama-3.3-70b-instruct/benchmarks","date":"2026-03-11"},{"name":"SciCode","score":26,"metric":"percent correct","source":"https://openrouter.ai/meta-llama/llama-3.3-70b-instruct/benchmarks","date":"2026-03-11"},{"name":"Terminal-Bench Hard","score":3,"metric":"success rate","source":"https://openrouter.ai/meta-llama/llama-3.3-70b-instruct/benchmarks","date":"2026-03-11"}]},"meta/llama-3.2-3b-instruct":{"id":"meta/llama-3.2-3b-instruct","name":"Llama 3.2 3B Instruct","description":"Small instruction-tuned Llama model for multilingual dialogue and agentic retrieval","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"license":"Llama 3.2 Community License","weights":[{"label":"Hugging Face","url":"https://huggingface.co/meta-llama/Llama-3.2-3B-Instruct"}]},"meta/llama-4-maverick-17b-instruct":{"id":"meta/llama-4-maverick-17b-instruct","name":"Llama 4 Maverick 17B Instruct","description":"Open multimodal Llama for strong reasoning with efficient everyday serving","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":16384},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/meta-llama/Llama-4-Maverick-17B-128E-Instruct"}],"benchmarks":[{"name":"Aider Polyglot","score":15.6,"metric":"percent correct","source":"https://aider.chat/docs/leaderboards/","date":"2025-04-06"},{"name":"SWE-Bench Pro","score":5.24,"metric":"resolve rate","dataset":"public","source":"https://labs.scale.com/leaderboard/swe_bench_pro_public"}]},"meta/muse-wonder-1.0":{"id":"meta/muse-wonder-1.0","name":"Muse Wonder","description":"Muse Wonder is a frontier multimodal reasoning model from Meta for personal and professional agents, long-horizon coding, and computer use. It combines Active Memory, which carries relevant context across conversations, with Active Focus, which keeps it working through blockers toward a goal.","family":"muse","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-10-12","last_updated":"2026-10-12","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"links":[{"label":"Announcement","url":"https://research.meta.ai/blog/muse-wonder","type":"announcement"},{"label":"Model card","url":"https://research.meta.ai/static/muse-wonder-model-card","type":"model_card"}],"benchmarks":[{"name":"Harvey Legal Agent Benchmark","score":10,"metric":"all-pass rate","variant":"max effort; Artificial Analysis LAB-AA","version":"1.1","source":"https://research.meta.ai/static/muse-wonder-model-card","date":"2026-10-12"},{"name":"Finance Agent","score":56.6,"metric":"accuracy","variant":"max effort","dataset":"Vals AI private test set","version":"v2","source":"https://research.meta.ai/static/muse-wonder-model-card","date":"2026-10-12"},{"name":"HealthBench Professional","score":60.5,"metric":"length-normalized weighted rubric score","variant":"max effort","source":"https://research.meta.ai/static/muse-wonder-model-card","date":"2026-10-12"},{"name":"AutomationBench","score":39.1,"metric":"strict pass@1","variant":"max effort; Artificial Analysis AutomationBench-AA","source":"https://research.meta.ai/static/muse-wonder-model-card","date":"2026-10-12"},{"name":"GDP.pdf","score":30.8,"metric":"score","variant":"max effort","source":"https://research.meta.ai/static/muse-wonder-model-card","date":"2026-10-12"},{"name":"JobBench","score":70.5,"metric":"mean reward","variant":"max effort","source":"https://research.meta.ai/static/muse-wonder-model-card","date":"2026-10-12"},{"name":"OSWorld","score":72.6,"metric":"mean partial score","variant":"max effort","version":"2.1","source":"https://research.meta.ai/static/muse-wonder-model-card","date":"2026-10-12"},{"name":"OSWorld","score":34.3,"metric":"binary completion rate","variant":"max effort","version":"2.1","source":"https://research.meta.ai/static/muse-wonder-model-card","date":"2026-10-12"},{"name":"Terminal-Bench","score":52,"metric":"accuracy","variant":"max effort; Artificial Analysis Terminal-Bench 4.0-AA","version":"4.0","source":"https://research.meta.ai/static/muse-wonder-model-card","date":"2026-10-12"},{"name":"DeepSWE","score":78.8,"metric":"pass@1","variant":"max effort","version":"1.1","source":"https://research.meta.ai/static/muse-wonder-model-card","date":"2026-10-12"},{"name":"SWE-Marathon","score":53.1,"metric":"pass@1","variant":"max effort","source":"https://research.meta.ai/static/muse-wonder-model-card","date":"2026-10-12"},{"name":"ProgramBench","score":50,"metric":"share of rollouts passing at least 95% of tests","harness":"mini-swe-agent","variant":"max effort","source":"https://research.meta.ai/static/muse-wonder-model-card","date":"2026-10-12"},{"name":"Artificial Analysis Intelligence Index","score":52,"metric":"index score","variant":"max effort","version":"4.3","source":"https://research.meta.ai/static/muse-wonder-model-card","date":"2026-10-12"}]},"meta/muse-glimmer-30b":{"id":"meta/muse-glimmer-30b","name":"Muse Glimmer 30B","description":"Muse Glimmer is a 30-billion-parameter open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark for always-on local agents, tool use, coding, and image understanding.","family":"muse","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-08-10","last_updated":"2026-08-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"license":"Apache 2.0","links":[{"label":"Announcement","url":"https://research.meta.ai/blog/introducing-muse-glimmer-open-agentic-model","type":"announcement"},{"label":"Model card","url":"https://huggingface.co/meta-models/Muse-Glimmer-30B","type":"model_card"},{"label":"Developer docs","url":"https://developer.meta.com/ai/models/muse-glimmer/","type":"docs"}],"weights":[{"label":"Hugging Face","url":"https://huggingface.co/meta-models/Muse-Glimmer-30B"}],"benchmarks":[{"name":"MCP Atlas","score":75.5,"metric":"success rate","variant":"public","source":"https://huggingface.co/meta-models/Muse-Glimmer-30B","date":"2026-08-10"},{"name":"DeepSearch QA","score":74.6,"source":"https://huggingface.co/meta-models/Muse-Glimmer-30B","date":"2026-08-10"},{"name":"SWE-Bench Pro","score":51.2,"metric":"resolve rate","source":"https://huggingface.co/meta-models/Muse-Glimmer-30B","date":"2026-08-10"},{"name":"SWE-Bench Verified","score":76,"metric":"resolve rate","source":"https://huggingface.co/meta-models/Muse-Glimmer-30B","date":"2026-08-10"},{"name":"Terminal-Bench","score":51.7,"metric":"success rate","variant":"with terminus2","version":"2.1","source":"https://huggingface.co/meta-models/Muse-Glimmer-30B","date":"2026-08-10"},{"name":"OSWorld-Verified","score":65.9,"metric":"success rate","source":"https://huggingface.co/meta-models/Muse-Glimmer-30B","date":"2026-08-10"},{"name":"AIME 2026","score":94.7,"metric":"accuracy","source":"https://huggingface.co/meta-models/Muse-Glimmer-30B","date":"2026-08-10"},{"name":"GPQA Diamond","score":83.5,"metric":"accuracy","variant":"AA","source":"https://huggingface.co/meta-models/Muse-Glimmer-30B","date":"2026-08-10"},{"name":"CharXiv Reasoning","score":78.8,"metric":"accuracy","source":"https://huggingface.co/meta-models/Muse-Glimmer-30B","date":"2026-08-10"}]},"meta/llama-3.2-1b-instruct":{"id":"meta/llama-3.2-1b-instruct","name":"Llama 3.2 1B Instruct","description":"Compact instruction-tuned Llama model for multilingual dialogue and on-device assistants","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32000},"license":"Llama 3.2 Community License","weights":[{"label":"Hugging Face","url":"https://huggingface.co/meta-llama/Llama-3.2-1B-Instruct"}]},"meta/llama-3.2-90b-vision-instruct":{"id":"meta/llama-3.2-90b-vision-instruct","name":"Llama-3.2-90B-Vision-Instruct","description":"Large open multimodal Llama model for image reasoning, document QA, and captioning","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"license":"Llama 3.2 Community License","weights":[{"label":"Hugging Face","url":"https://huggingface.co/meta-llama/Llama-3.2-90B-Vision-Instruct"}]},"meta/muse-wonder-full":{"id":"meta/muse-wonder-full","name":"Muse Wonder Full","description":"Muse Wonder Full is the most capable model in Meta's Muse Wonder series, built for long-running professional agents and ultra-long-horizon software engineering. It shares Muse Wonder's Active Memory and Active Focus capabilities and generally scores higher on Meta's agentic and coding evaluations.","family":"muse","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-10-12","last_updated":"2026-10-12","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576},"links":[{"label":"Announcement","url":"https://research.meta.ai/blog/muse-wonder","type":"announcement"},{"label":"Model card","url":"https://research.meta.ai/static/muse-wonder-model-card","type":"model_card"}],"benchmarks":[{"name":"Harvey Legal Agent Benchmark","score":13.1,"metric":"all-pass rate","variant":"max effort; Artificial Analysis LAB-AA","version":"1.1","source":"https://research.meta.ai/static/muse-wonder-model-card","date":"2026-10-12"},{"name":"Finance Agent","score":59.1,"metric":"accuracy","variant":"max effort","dataset":"Vals AI private test set","version":"v2","source":"https://research.meta.ai/static/muse-wonder-model-card","date":"2026-10-12"},{"name":"HealthBench Professional","score":62.3,"metric":"length-normalized weighted rubric score","variant":"max effort","source":"https://research.meta.ai/static/muse-wonder-model-card","date":"2026-10-12"},{"name":"AutomationBench","score":45.4,"metric":"strict pass@1","variant":"max effort; Artificial Analysis AutomationBench-AA","source":"https://research.meta.ai/static/muse-wonder-model-card","date":"2026-10-12"},{"name":"GDP.pdf","score":33.2,"metric":"score","variant":"max effort","source":"https://research.meta.ai/static/muse-wonder-model-card","date":"2026-10-12"},{"name":"JobBench","score":75.3,"metric":"mean reward","variant":"max effort","source":"https://research.meta.ai/static/muse-wonder-model-card","date":"2026-10-12"},{"name":"OSWorld","score":78.4,"metric":"mean partial score","variant":"max effort","version":"2.1","source":"https://research.meta.ai/static/muse-wonder-model-card","date":"2026-10-12"},{"name":"OSWorld","score":39.4,"metric":"binary completion rate","variant":"max effort","version":"2.1","source":"https://research.meta.ai/static/muse-wonder-model-card","date":"2026-10-12"},{"name":"Terminal-Bench","score":57.1,"metric":"accuracy","variant":"max effort; Artificial Analysis Terminal-Bench 4.0-AA","version":"4.0","source":"https://research.meta.ai/static/muse-wonder-model-card","date":"2026-10-12"},{"name":"DeepSWE","score":79,"metric":"pass@1","variant":"max effort","version":"1.1","source":"https://research.meta.ai/static/muse-wonder-model-card","date":"2026-10-12"},{"name":"SWE-Marathon","score":57.5,"metric":"pass@1","variant":"xhigh effort","source":"https://research.meta.ai/static/muse-wonder-model-card","date":"2026-10-12"},{"name":"ProgramBench","score":59,"metric":"share of rollouts passing at least 95% of tests","harness":"mini-swe-agent","variant":"max effort","source":"https://research.meta.ai/static/muse-wonder-model-card","date":"2026-10-12"},{"name":"Artificial Analysis Intelligence Index","score":56,"metric":"index score","variant":"max effort","version":"4.3","source":"https://research.meta.ai/static/muse-wonder-model-card","date":"2026-10-12"}]},"meta/llama-3.2-1b":{"id":"meta/llama-3.2-1b","name":"Llama-3.2-1B","description":"Compact open Llama base model for lightweight and on-device use","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"license":"Llama 3.2 Community License","weights":[{"label":"Hugging Face","url":"https://huggingface.co/meta-llama/Llama-3.2-1B"}]},"meta/llama-prompt-guard-2-22m":{"id":"meta/llama-prompt-guard-2-22m","name":"Llama Prompt Guard 2 22M","description":"22M-parameter (DeBERTa-xsmall) English classifier that detects prompt injection and jailbreak attempts","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":512,"output":512},"license":"Llama 4 Community License","weights":[{"label":"Hugging Face","url":"https://huggingface.co/meta-llama/Llama-Prompt-Guard-2-22M"}]},"meta/muse-spark-1.3":{"id":"meta/muse-spark-1.3","name":"Muse Spark 1.3","description":"Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It improves long-horizon agent collaboration, instruction following, and coding efficiency relative to Muse Spark 1.2.","family":"muse","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"benchmarks":[{"name":"GDPval-AA","score":1754,"metric":"Elo","harness":"Stirrup","variant":"max effort","version":"2","source":"https://research.meta.ai/blog/introducing-muse-spark-1-3","date":"2026-09-02"},{"name":"JobBench","score":64.9,"metric":"mean rubric score","harness":"OpenCode","variant":"max effort","source":"https://research.meta.ai/blog/introducing-muse-spark-1-3","date":"2026-09-02"},{"name":"OSWorld","score":66.9,"metric":"mean partial score","harness":"Meta internal evaluation framework","variant":"max effort","dataset":"08.08","version":"2.0","source":"https://research.meta.ai/blog/introducing-muse-spark-1-3","date":"2026-09-02"},{"name":"OSWorld","score":32,"metric":"binary completion rate","harness":"Meta internal evaluation framework","variant":"max effort","dataset":"08.08","version":"2.0","source":"https://research.meta.ai/blog/introducing-muse-spark-1-3","date":"2026-09-02"},{"name":"DeepSearchQA","score":90.3,"metric":"F1","harness":"Meta browser harness","variant":"max effort","source":"https://research.meta.ai/blog/introducing-muse-spark-1-3","date":"2026-09-02"},{"name":"Agentic IF Index","score":57.8,"metric":"score","variant":"max effort","dataset":"Meta internal","source":"https://research.meta.ai/blog/introducing-muse-spark-1-3","date":"2026-09-02"},{"name":"AutomationBench","score":49.6,"metric":"pass@1","variant":"max effort","dataset":"public v3 task set","source":"https://research.meta.ai/blog/introducing-muse-spark-1-3","date":"2026-09-02"},{"name":"MRCR","score":98.5,"metric":"mean sequence-match ratio","variant":"max effort","dataset":"8-needle, 256K-512K","version":"2","source":"https://research.meta.ai/blog/introducing-muse-spark-1-3","date":"2026-09-02"},{"name":"MRCR","score":98.1,"metric":"mean sequence-match ratio","variant":"max effort","dataset":"8-needle, 512K-1M","version":"2","source":"https://research.meta.ai/blog/introducing-muse-spark-1-3","date":"2026-09-02"},{"name":"DeepSWE","score":75.4,"metric":"pass@1","harness":"mini-swe-agent","variant":"max effort","version":"1.1","source":"https://research.meta.ai/blog/introducing-muse-spark-1-3","date":"2026-09-02"},{"name":"SWE-Atlas Codebase QnA","score":59.4,"metric":"pass@1","harness":"mini-swe-agent","variant":"max effort","source":"https://research.meta.ai/blog/introducing-muse-spark-1-3","date":"2026-09-02"},{"name":"Terminal-Bench","score":88.8,"metric":"pass@1","harness":"Muse Code","variant":"max effort","version":"2.1","source":"https://research.meta.ai/blog/introducing-muse-spark-1-3","date":"2026-09-02"}]},"meta/llama-3.1-8b-instruct":{"id":"meta/llama-3.1-8b-instruct","name":"Llama-3.1-8B-Instruct","description":"Compact open Llama model for lightweight chat, drafting, and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/meta-llama/Llama-3.1-8B-Instruct"}]},"meta/llama-guard-4-12b":{"id":"meta/llama-guard-4-12b","name":"Llama Guard 4 12B","description":"Multimodal Llama 4-based safety classifier for moderating prompts and responses","family":"llama","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":163840},"license":"Llama 4 Community License","weights":[{"label":"Hugging Face","url":"https://huggingface.co/meta-llama/Llama-Guard-4-12B"}]},"meta/llama-3.2-11b-vision-instruct":{"id":"meta/llama-3.2-11b-vision-instruct","name":"Llama-3.2-11B-Vision-Instruct","description":"Open multimodal Llama model for image understanding, captioning, and visual QA","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/meta-llama/Llama-3.2-11B-Vision-Instruct"}]},"meta/llama-prompt-guard-2-86m":{"id":"meta/llama-prompt-guard-2-86m","name":"Llama Prompt Guard 2 86M","description":"86M-parameter (mDeBERTa-base) multilingual classifier that detects prompt injection and jailbreak attempts","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":512,"output":512},"license":"Llama 4 Community License","weights":[{"label":"Hugging Face","url":"https://huggingface.co/meta-llama/Llama-Prompt-Guard-2-86M"}]},"meta/llama-3.2-3b":{"id":"meta/llama-3.2-3b","name":"Llama-3.2-3B","description":"Small open Llama base model for lightweight text generation and self-hosting","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-09-25","last_updated":"2024-09-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"license":"Llama 3.2 Community License","weights":[{"label":"Hugging Face","url":"https://huggingface.co/meta-llama/Llama-3.2-3B"}]},"meta/llama-3.1-70b-instruct":{"id":"meta/llama-3.1-70b-instruct","name":"Llama-3.1-70B-Instruct","description":"Open Llama instruction model for multilingual chat, reasoning, and coding","family":"llama","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/meta-llama/Llama-3.1-70B-Instruct"}]},"meta/muse-spark-open":{"id":"meta/muse-spark-open","name":"Muse Spark Open","description":"Muse Spark Open is an 816B total / 42B active open-weight multimodal reasoning model from Meta for coding and long-horizon agentic work, with 1M-token context.","family":"muse","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2026-01-04","release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"license":"Muse Model License","links":[{"label":"Announcement","url":"https://research.meta.ai/blog/muse-spark-open","type":"announcement"},{"label":"Model card","url":"https://huggingface.co/meta-models/Muse-Spark-Open-816B-A42B","type":"model_card"},{"label":"Developer docs","url":"https://dev.meta.ai/docs/muse-spark-open","type":"docs"},{"label":"License","url":"https://huggingface.co/meta-models/Muse-Spark-Open-816B-A42B/blob/main/LICENSE.md","type":"license"}],"weights":[{"label":"Hugging Face (BF16)","url":"https://huggingface.co/meta-models/Muse-Spark-Open-816B-A42B"},{"label":"Hugging Face (NVFP4)","url":"https://huggingface.co/meta-models/Muse-Spark-Open-816B-A42B-NVFP4"}],"benchmarks":[{"name":"Artificial Analysis Intelligence Index","score":39,"metric":"index score","variant":"xhigh effort","source":"https://huggingface.co/meta-models/Muse-Spark-Open-816B-A42B","date":"2026-10-15"},{"name":"JobBench","score":61.8,"variant":"xhigh effort","source":"https://huggingface.co/meta-models/Muse-Spark-Open-816B-A42B","date":"2026-10-15"},{"name":"MCP Atlas","score":89.2,"variant":"xhigh effort","source":"https://huggingface.co/meta-models/Muse-Spark-Open-816B-A42B","date":"2026-10-15"},{"name":"GDPval-AA","score":1401,"metric":"Elo","variant":"xhigh effort","version":"2","source":"https://huggingface.co/meta-models/Muse-Spark-Open-816B-A42B","date":"2026-10-15"},{"name":"OSWorld-Verified","score":80.4,"variant":"xhigh effort","source":"https://huggingface.co/meta-models/Muse-Spark-Open-816B-A42B","date":"2026-10-15"},{"name":"Toolathlon-Verified","score":72.8,"variant":"xhigh effort","source":"https://huggingface.co/meta-models/Muse-Spark-Open-816B-A42B","date":"2026-10-15"},{"name":"HLE-Diamond","score":26.1,"variant":"xhigh effort, no tools","source":"https://huggingface.co/meta-models/Muse-Spark-Open-816B-A42B","date":"2026-10-15"},{"name":"AA-Omniscience","score":25,"variant":"xhigh effort","source":"https://huggingface.co/meta-models/Muse-Spark-Open-816B-A42B","date":"2026-10-15"},{"name":"SciCode","score":58.5,"variant":"xhigh effort","source":"https://huggingface.co/meta-models/Muse-Spark-Open-816B-A42B","date":"2026-10-15"},{"name":"Humanity's Last Exam","score":61,"variant":"xhigh effort, with tools","source":"https://huggingface.co/meta-models/Muse-Spark-Open-816B-A42B","date":"2026-10-15"},{"name":"GPQA Diamond","score":89,"variant":"xhigh effort","source":"https://huggingface.co/meta-models/Muse-Spark-Open-816B-A42B","date":"2026-10-15"},{"name":"DeepSWE","score":60,"harness":"preferred harness","variant":"xhigh effort","version":"1.1","source":"https://huggingface.co/meta-models/Muse-Spark-Open-816B-A42B","date":"2026-10-15"},{"name":"SWE-Atlas Codebase QnA","score":58,"harness":"preferred harness","variant":"xhigh effort","source":"https://huggingface.co/meta-models/Muse-Spark-Open-816B-A42B","date":"2026-10-15"},{"name":"Terminal-Bench","score":76.4,"harness":"Terminus 2","variant":"xhigh effort","version":"2.1","source":"https://huggingface.co/meta-models/Muse-Spark-Open-816B-A42B","date":"2026-10-15"},{"name":"ZeroBench","score":50,"variant":"xhigh effort, with tools","source":"https://huggingface.co/meta-models/Muse-Spark-Open-816B-A42B","date":"2026-10-15"},{"name":"BabyVision","score":80.2,"variant":"xhigh effort, with tools","source":"https://huggingface.co/meta-models/Muse-Spark-Open-816B-A42B","date":"2026-10-15"},{"name":"SimpleVQA","score":73.6,"variant":"xhigh effort, with tools","source":"https://huggingface.co/meta-models/Muse-Spark-Open-816B-A42B","date":"2026-10-15"},{"name":"WorldVQA","score":49.4,"variant":"xhigh effort, with tools","source":"https://huggingface.co/meta-models/Muse-Spark-Open-816B-A42B","date":"2026-10-15"},{"name":"IntPhys2","score":68.3,"variant":"xhigh effort","source":"https://huggingface.co/meta-models/Muse-Spark-Open-816B-A42B","date":"2026-10-15"},{"name":"Video-MME","score":61.5,"variant":"xhigh effort","version":"2","source":"https://huggingface.co/meta-models/Muse-Spark-Open-816B-A42B","date":"2026-10-15"}]},"meta/muse-spark-1.2":{"id":"meta/muse-spark-1.2","name":"Muse Spark 1.2","description":"Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.","family":"muse","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-08-05","last_updated":"2026-08-05","modalities":{"input":["text","image","video","pdf","audio"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"benchmarks":[{"name":"GDPval-AA","score":1615,"metric":"Elo","harness":"Stirrup","variant":"xhigh effort","version":"2","source":"https://research.meta.ai/blog/introducing-muse-spark-1-3","date":"2026-09-02"},{"name":"JobBench","score":61.6,"metric":"mean rubric score","harness":"OpenCode","variant":"xhigh effort","source":"https://research.meta.ai/blog/introducing-muse-spark-1-3","date":"2026-09-02"},{"name":"OSWorld","score":47.6,"metric":"mean partial score","harness":"Meta internal evaluation framework","variant":"xhigh effort","dataset":"06.24","version":"2.0","source":"https://research.meta.ai/blog/introducing-muse-spark-1-3","date":"2026-09-02"},{"name":"OSWorld","score":17.9,"metric":"binary completion rate","harness":"Meta internal evaluation framework","variant":"xhigh effort","dataset":"06.24","version":"2.0","source":"https://research.meta.ai/blog/introducing-muse-spark-1-3","date":"2026-09-02"},{"name":"DeepSearchQA","score":85.9,"metric":"F1","harness":"Meta browser harness","variant":"xhigh effort","source":"https://research.meta.ai/blog/introducing-muse-spark-1-3","date":"2026-09-02"},{"name":"Agentic IF Index","score":46.2,"metric":"score","variant":"xhigh effort","dataset":"Meta internal","source":"https://research.meta.ai/blog/introducing-muse-spark-1-3","date":"2026-09-02"},{"name":"AutomationBench","score":38.2,"metric":"pass@1","variant":"xhigh effort","dataset":"public v3 task set","source":"https://research.meta.ai/blog/introducing-muse-spark-1-3","date":"2026-09-02"},{"name":"MRCR","score":66.3,"metric":"mean sequence-match ratio","variant":"xhigh effort","dataset":"8-needle, 256K-512K","version":"2","source":"https://research.meta.ai/blog/introducing-muse-spark-1-3","date":"2026-09-02"},{"name":"MRCR","score":55.5,"metric":"mean sequence-match ratio","variant":"xhigh effort","dataset":"8-needle, 512K-1M","version":"2","source":"https://research.meta.ai/blog/introducing-muse-spark-1-3","date":"2026-09-02"},{"name":"DeepSWE","score":55,"metric":"pass@1","harness":"mini-swe-agent","variant":"xhigh effort","version":"1.1","source":"https://research.meta.ai/blog/introducing-muse-spark-1-3","date":"2026-09-02"},{"name":"SWE-Atlas Codebase QnA","score":46.2,"metric":"pass@1","harness":"mini-swe-agent","variant":"xhigh effort","source":"https://research.meta.ai/blog/introducing-muse-spark-1-3","date":"2026-09-02"},{"name":"Terminal-Bench","score":82.9,"metric":"pass@1","harness":"Muse Code","variant":"xhigh effort","version":"2.1","source":"https://research.meta.ai/blog/introducing-muse-spark-1-3","date":"2026-09-02"},{"name":"DeepSWE","score":59.3,"metric":"pass@1","harness":"Muse Code","variant":"xhigh effort","version":"1.1","source":"https://research.meta.ai/blog/introducing-muse-code-and-muse-spark-1-2","date":"2026-08-05"},{"name":"MCP Atlas","score":90.3,"metric":"pass rate","harness":"Scale AI","variant":"xhigh effort","source":"https://research.meta.ai/blog/introducing-muse-code-and-muse-spark-1-2","date":"2026-08-05"}]},"meta/llama-4-scout-17b-instruct":{"id":"meta/llama-4-scout-17b-instruct","name":"Llama 4 Scout 17B Instruct","description":"Open Llama with long-context vision for efficient multimodal agents","family":"llama","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-04-05","last_updated":"2025-04-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":10000000,"output":16384},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/meta-llama/Llama-4-Scout-17B-16E-Instruct"}]},"meta/llama-guard-3-8b":{"id":"meta/llama-guard-3-8b","name":"Llama-Guard-3-8B","description":"Llama 3.1-based safety classifier for moderating prompts and model responses","family":"llama","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2023-12","release_date":"2024-07-23","last_updated":"2024-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/meta-llama/Llama-Guard-3-8B"}]},"meta/muse-spark-1.1":{"id":"meta/muse-spark-1.1","name":"Muse Spark 1.1","description":"Muse Spark is a natively multimodal reasoning model with support for tool-use, visual chain of thought, and multi-agent orchestration.","family":"muse","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-04-08","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072},"benchmarks":[{"name":"SWE-Bench Pro","score":61.5,"metric":"resolve rate","source":"https://ai.meta.com/blog/introducing-muse-spark-meta-model-api/","date":"2026-07-09"},{"name":"Terminal-Bench","score":80,"metric":"success rate","version":"2.1","source":"https://ai.meta.com/blog/introducing-muse-spark-meta-model-api/","date":"2026-07-09"},{"name":"DeepSWE","score":53.3,"metric":"resolve rate","version":"1.1","source":"https://ai.meta.com/blog/introducing-muse-spark-meta-model-api/","date":"2026-07-09"},{"name":"MCP Atlas","score":88.1,"metric":"success rate","source":"https://ai.meta.com/blog/introducing-muse-spark-meta-model-api/","date":"2026-07-09"},{"name":"JobBench","score":54.7,"metric":"success rate","source":"https://ai.meta.com/blog/introducing-muse-spark-meta-model-api/","date":"2026-07-09"},{"name":"Toolathlon-Verified","score":75.6,"metric":"success rate","source":"https://ai.meta.com/blog/introducing-muse-spark-meta-model-api/","date":"2026-07-09"},{"name":"Humanity's Last Exam","score":62.1,"metric":"accuracy","variant":"with tools","source":"https://ai.meta.com/blog/introducing-muse-spark-meta-model-api/","date":"2026-07-09"},{"name":"OSWorld-Verified","score":80.8,"metric":"success rate","source":"https://ai.meta.com/blog/introducing-muse-spark-meta-model-api/","date":"2026-07-09"},{"name":"Finance Agent","score":57.2,"metric":"accuracy","version":"v2","source":"https://ai.meta.com/blog/introducing-muse-spark-meta-model-api/","date":"2026-07-09"},{"name":"CharXiv Reasoning","score":88.4,"metric":"accuracy","source":"https://ai.meta.com/blog/introducing-muse-spark-meta-model-api/","date":"2026-07-09"},{"name":"BabyVision","score":76.3,"metric":"accuracy","source":"https://ai.meta.com/blog/introducing-muse-spark-meta-model-api/","date":"2026-07-09"}]},"morph/morph-v3-large":{"id":"morph/morph-v3-large","name":"Morph V3 Large","description":"Flagship model for demanding analysis, coding, and production agent workflows","family":"morph","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-07-07","last_updated":"2025-07-07","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072}},"morph/morph-v3-fast":{"id":"morph/morph-v3-fast","name":"Morph V3 Fast","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"morph","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-07-07","last_updated":"2025-07-07","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072}},"minimax/MiniMax-Text-01":{"id":"minimax/MiniMax-Text-01","name":"MiniMax Text 01 (MiniMax)","description":"MiniMax model for chat, coding, office work, and agentic tasks","family":"minimax","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-01-15","last_updated":"2025-01-15","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072}},"minimax/MiniMax-M2-Her":{"id":"minimax/MiniMax-M2-Her","name":"MiniMax-M2 Her","description":"MiniMax M2 variant tuned for conversational and character-driven agent interactions","family":"minimax","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-01-23","last_updated":"2026-01-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":65536,"output":2048}},"minimax/MiniMax-M2.1":{"id":"minimax/MiniMax-M2.1","name":"MiniMax-M2.1","description":"Earlier MiniMax agent model for practical coding and productivity tasks","family":"minimax","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2025-12-23","last_updated":"2025-12-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/MiniMaxAI/MiniMax-M2.1"}],"benchmarks":[{"name":"SWE-Bench Verified","score":74,"metric":"resolved","source":"https://huggingface.co/MiniMaxAI/MiniMax-M2.1"},{"name":"SWE-Bench Pro","score":36.81,"metric":"resolve rate","dataset":"public","source":"https://labs.scale.com/leaderboard/swe_bench_pro_public"}]},"minimax/MiniMax-M3.1-Flash":{"id":"minimax/MiniMax-M3.1-Flash","name":"MiniMax-M3.1-Flash","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-10-19","last_updated":"2026-10-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":512000}},"minimax/MiniMax-M2.5-highspeed":{"id":"minimax/MiniMax-M2.5-highspeed","name":"MiniMax-M2.5-highspeed","description":"High-speed MiniMax model for low-latency coding and agent workflows","family":"minimax","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-02-13","last_updated":"2026-02-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/MiniMaxAI/MiniMax-M2.5"}]},"minimax/MiniMax-M2.5":{"id":"minimax/MiniMax-M2.5","name":"MiniMax-M2.5","description":"Prior MiniMax coding model for agent workflows, office edits, and automation","family":"minimax","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-02-12","last_updated":"2026-02-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/MiniMaxAI/MiniMax-M2.5"}],"benchmarks":[{"name":"SWE-Bench Verified","score":75.8,"metric":"resolved","source":"https://www.swebench.com/"},{"name":"SWE-Atlas Codebase QnA","score":10.3,"metric":"score","harness":"Mini-SWE-Agent","source":"https://labs.scale.com/leaderboard/sweatlas-qna"},{"name":"SWE-Atlas Refactoring","score":19.52,"metric":"score","harness":"Mini-SWE-Agent","source":"https://labs.scale.com/leaderboard/sweatlas-refactoring"},{"name":"SWE-Atlas Test Writing","score":18.6,"metric":"score","harness":"Mini-SWE-Agent","source":"https://labs.scale.com/leaderboard/sweatlas-tw"}]},"minimax/MiniMax-M3.1-Flash-Preview":{"id":"minimax/MiniMax-M3.1-Flash-Preview","name":"MiniMax-M3.1-Flash-Preview","description":"Frontier multimodal coding model with 1M context window and tunable thinking depth","family":"minimax","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-09-27","last_updated":"2026-09-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":512000}},"minimax/MiniMax-H3":{"id":"minimax/MiniMax-H3","type":"video","name":"MiniMax H3","description":"MiniMax general-purpose video model with multimodal references and synchronized stereo audio. Open H3-Base checkpoints; Context-IR and 2K regeneration are not included in the initial weights release.","family":"minimax","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-07-31","last_updated":"2026-10-04","modalities":{"input":["text","image","video","audio"],"output":["video","audio"]},"open_weights":true,"limit":{"context":0,"output":0},"license":"MiniMax H3 Community License Agreement","links":[{"url":"https://www.minimax.io/blog/minimax-h3","type":"announcement"},{"url":"https://www.minimax.io/news/minimax-h3-open-source","type":"announcement"},{"url":"https://huggingface.co/MiniMaxAI/MiniMax-H3","type":"model_card"},{"url":"https://huggingface.co/MiniMaxAI/MiniMax-H3/blob/main/LICENSE","type":"license"}],"weights":[{"label":"Hugging Face: H3-Base FL2VA / Ref2VA","url":"https://huggingface.co/MiniMaxAI/MiniMax-H3","format":"safetensors"}]},"minimax/image-01":{"id":"minimax/image-01","type":"image","name":"MiniMax image-01","description":"MiniMax text-to-image generation model with reference-image support","family":"minimax","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-02-15","last_updated":"2026-08-25","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0}},"minimax/minimax-m1-80k":{"id":"minimax/minimax-m1-80k","name":"MiniMax M1 80K","description":"Open hybrid-attention MoE reasoning model with a million-token context and an 80K thinking budget","family":"minimax","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2025-06-17","last_updated":"2025-06-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":80000},"license":"Apache-2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/MiniMaxAI/MiniMax-M1-80k"}]},"minimax/MiniMax-M2.7":{"id":"minimax/MiniMax-M2.7","name":"MiniMax-M2.7","description":"Open MiniMax flagship for coding agents, office automation, and complex environments","family":"minimax","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/MiniMaxAI/MiniMax-M2.7"}],"benchmarks":[{"name":"SWE-Bench Verified","score":79.9,"metric":"resolved","harness":"Claude Code","source":"https://www.minimax.io/blog/minimax-m3","date":"2026-06-01"},{"name":"SWE-Bench Pro","score":56.2,"metric":"resolve rate","harness":"Claude Code","source":"https://www.minimax.io/blog/minimax-m3","date":"2026-06-01"},{"name":"Terminal-Bench","score":51.1,"metric":"success rate","version":"2.1","source":"https://www.minimax.io/blog/minimax-m3","date":"2026-06-01"}]},"minimax/MiniMax-M2":{"id":"minimax/MiniMax-M2","name":"MiniMax-M2","description":"Efficient open MiniMax model built for coding agents and tool-heavy workflows","family":"minimax","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2025-10-27","last_updated":"2025-10-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/MiniMaxAI/MiniMax-M2"}],"benchmarks":[{"name":"SWE-Bench Verified","score":69.4,"metric":"resolved","source":"https://huggingface.co/MiniMaxAI/MiniMax-M2"}]},"minimax/MiniMax-M2.7-highspeed":{"id":"minimax/MiniMax-M2.7-highspeed","name":"MiniMax-M2.7-highspeed","description":"Low-latency M2.7 variant for interactive coding plans and agent loops","family":"minimax","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":204800,"output":131072},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/MiniMaxAI/MiniMax-M2.7"}]},"minimax/MiniMax-M3":{"id":"minimax/MiniMax-M3","name":"MiniMax-M3","description":"MiniMax multimodal model for long-context coding, perception, and agent planning","family":"minimax","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-06-01","last_updated":"2026-06-01","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":512000},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/MiniMaxAI/MiniMax-M3"}],"benchmarks":[{"name":"SWE-Bench Verified","score":80.5,"metric":"resolved","harness":"Claude Code","source":"https://www.minimax.io/blog/minimax-m3","date":"2026-06-01"},{"name":"SWE-Bench Pro","score":59,"metric":"resolve rate","harness":"Claude Code","source":"https://www.minimax.io/blog/minimax-m3","date":"2026-06-01"},{"name":"Terminal-Bench","score":66,"metric":"success rate","version":"2.1","source":"https://www.minimax.io/blog/minimax-m3","date":"2026-06-01"},{"name":"BrowseComp","score":83.52,"metric":"accuracy","source":"https://www.minimax.io/blog/minimax-m3","date":"2026-06-01"},{"name":"MCP Atlas","score":74.2,"metric":"success rate","source":"https://www.minimax.io/blog/minimax-m3","date":"2026-06-01"},{"name":"OSWorld-Verified","score":70.06,"metric":"success rate","source":"https://www.minimax.io/blog/minimax-m3","date":"2026-06-01"}]},"sakana/fugu-ultra-v2":{"id":"sakana/fugu-ultra-v2","name":"Fugu Ultra v2","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2026-08-28","release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000}},"sakana/fugu":{"id":"sakana/fugu","name":"Fugu","description":"Multi-agent model for routing expert agents across complex analytical tasks","family":"fugu","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000},"links":[{"label":"Official model catalog","url":"https://raw.githubusercontent.com/SakanaAI/fugu/refs/heads/main/configs/files/fugu.json","type":"docs"}],"benchmarks":[{"name":"SWE Bench Pro","score":59,"source":"https://console.sakana.ai/models"},{"name":"Terminal Bench 2.1","score":80.2,"source":"https://console.sakana.ai/models"},{"name":"LiveCodeBench","score":92.9,"source":"https://console.sakana.ai/models"},{"name":"LiveCodeBench Pro","score":87.8,"source":"https://console.sakana.ai/models"},{"name":"Humanity’s Last Exam","score":47.2,"source":"https://console.sakana.ai/models"},{"name":"CharXiv Reasoning","score":85.1,"source":"https://console.sakana.ai/models"},{"name":"GPQA Diamond","score":95.5,"source":"https://console.sakana.ai/models"},{"name":"SciCode","score":60.1,"source":"https://console.sakana.ai/models"},{"name":"τ3 Banking","score":21.7,"source":"https://console.sakana.ai/models"},{"name":"Long Context Reasoning","score":74.7,"source":"https://console.sakana.ai/models"},{"name":"MRCRv2","score":86.6,"source":"https://console.sakana.ai/models"},{"name":"CTI-REALM","score":67.5,"source":"https://console.sakana.ai/models"}]},"sakana/fugu-ultra":{"id":"sakana/fugu-ultra","name":"Fugu Ultra","description":"Quality-first multi-agent model for hard research, analysis, and competitions","family":"fugu","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"release_date":"2026-06-15","last_updated":"2026-06-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000},"links":[{"label":"Official model catalog","url":"https://raw.githubusercontent.com/SakanaAI/fugu/refs/heads/main/configs/files/fugu.json","type":"docs"}],"benchmarks":[{"name":"SWE Bench Pro","score":73.7,"source":"https://console.sakana.ai/models"},{"name":"Terminal Bench 2.1","score":82.1,"source":"https://console.sakana.ai/models"},{"name":"LiveCodeBench","score":93.2,"source":"https://console.sakana.ai/models"},{"name":"LiveCodeBench Pro","score":90.8,"source":"https://console.sakana.ai/models"},{"name":"Humanity’s Last Exam","score":50,"source":"https://console.sakana.ai/models"},{"name":"CharXiv Reasoning","score":86.6,"source":"https://console.sakana.ai/models"},{"name":"GPQA Diamond","score":95.5,"source":"https://console.sakana.ai/models"},{"name":"SciCode","score":58.7,"source":"https://console.sakana.ai/models"},{"name":"τ3 Banking","score":20.6,"source":"https://console.sakana.ai/models"},{"name":"Long Context Reasoning","score":73.3,"source":"https://console.sakana.ai/models"},{"name":"MRCRv2","score":93.6,"source":"https://console.sakana.ai/models"},{"name":"CTI-REALM","score":69.4,"source":"https://console.sakana.ai/models"}]},"sakana/sakana-namazu":{"id":"sakana/sakana-namazu","name":"Sakana Namazu","description":"Japanese-specialized reasoning model based on Kimi K2.6 and tuned for Japanese language, culture, and business workflows","family":"sakana-namazu","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"links":[{"label":"Official product page","url":"https://sakana.ai/namazu/","type":"announcement"},{"label":"Official model documentation","url":"https://console.sakana.ai/models?model=sakana-namazu","type":"docs"}],"benchmarks":[{"name":"AIME26","score":96.67,"source":"https://console.sakana.ai/models?model=sakana-namazu"},{"name":"MMLU-Pro","score":90.33,"source":"https://console.sakana.ai/models?model=sakana-namazu"},{"name":"LiveCodeBench v6","score":90.33,"source":"https://console.sakana.ai/models?model=sakana-namazu"},{"name":"JFBench","score":37.4,"source":"https://console.sakana.ai/models?model=sakana-namazu"},{"name":"Translation","score":52.2,"source":"https://console.sakana.ai/models?model=sakana-namazu"},{"name":"FairPoliticsQA","score":56.3,"source":"https://console.sakana.ai/models?model=sakana-namazu"}]},"sakana/fugu-max":{"id":"sakana/fugu-max","name":"Fugu Max","description":"Multi-agent model for routing expert agents across complex analytical tasks","family":"fugu","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000}},"sarvam/sarvam-30b":{"id":"sarvam/sarvam-30b","name":"Sarvam 30B","description":"Efficient Indian-language reasoning model for chat, coding, and multilingual work","family":"sarvam","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-02-18","last_updated":"2026-02-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000}},"sarvam/sarvam-105b":{"id":"sarvam/sarvam-105b","name":"Sarvam 105B","description":"Flagship Indian-language reasoning model for enterprise multilingual applications","family":"sarvam","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-02-18","last_updated":"2026-02-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072}},"sarvam/sarvam-m":{"id":"sarvam/sarvam-m","name":"Sarvam-M","description":"Hybrid-reasoning multilingual model for Indian languages built on Mistral Small","family":"sarvam","attachment":false,"reasoning":true,"tool_call":false,"temperature":true,"release_date":"2025-05-20","last_updated":"2025-05-20","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"license":"Apache 2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/sarvamai/sarvam-m"}]},"unbiased/pareto":{"id":"unbiased/pareto","name":"Pareto","description":"Blended multimodal model that runs multiple language models in parallel and returns a synthesized answer","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-09-17","last_updated":"2026-09-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072}},"mancer/weaver":{"id":"mancer/weaver","name":"Weaver (alpha)","description":"General-purpose chat model for instruction following, writing, and analysis","family":"alpha","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2023-06-30","release_date":"2023-08-02","last_updated":"2023-08-02","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8000,"output":6000}},"gryphe/mythomax-l2-13b":{"id":"gryphe/mythomax-l2-13b","name":"MythoMax L2 13B","description":"Llama 2 merge combining MythoLogic and Huginn for creative writing and roleplay","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2023-08-10","last_updated":"2023-08-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":4096,"output":3200},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/Gryphe/MythoMax-L2-13b"}]},"nvidia/llama-3.3-nemotron-super-49b-v1":{"id":"nvidia/llama-3.3-nemotron-super-49b-v1","name":"Llama 3.3 Nemotron Super 49B v1","description":"Nemotron model for efficient reasoning, coding, and specialized AI agents","family":"nemotron","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2025-03-18","last_updated":"2025-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072}},"nvidia/nemotron-3.5-super-120b-a12b":{"id":"nvidia/nemotron-3.5-super-120b-a12b","name":"Nemotron 3.5 Super 120B A12B","description":"NVIDIA Nemotron 3.5 Super reasoning model (120B total, 12B active) served with text-only input","family":"nemotron","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072}},"nvidia/nemotron-content-safety-reasoning-4b":{"id":"nvidia/nemotron-content-safety-reasoning-4b","name":"Nemotron Content Safety Reasoning 4B","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"nemotron","attachment":false,"reasoning":true,"tool_call":false,"temperature":false,"release_date":"2026-01-22","last_updated":"2026-01-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096}},"nvidia/llama-nemotron-embed-1b-v2":{"id":"nvidia/llama-nemotron-embed-1b-v2","type":"embedding","name":"Llama Nemotron Embed 1B v2","description":"Multilingual text embedding model for retrieval and question answering pipelines","family":"nemotron","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-10-16","last_updated":"2025-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":2048},"license":"nvidia-open-model-license","weights":[{"label":"Hugging Face","url":"https://huggingface.co/nvidia/llama-nemotron-embed-1b-v2"}]},"nvidia/nemotron-nano-9b-v2":{"id":"nvidia/nemotron-nano-9b-v2","name":"Nemotron Nano 9B v2","description":"Compact Nemotron model for efficient reasoning and deployable AI agents","family":"nemotron","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2024-09","release_date":"2025-08-18","last_updated":"2025-08-18","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072}},"nvidia/nemotron-mini-4b-instruct":{"id":"nvidia/nemotron-mini-4b-instruct","name":"Nemotron Mini 4B Instruct","description":"Compact Nemotron model for efficient reasoning and deployable AI agents","family":"nemotron","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2024-08-21","last_updated":"2024-08-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192}},"nvidia/nemotron-3-super-120b-a12b":{"id":"nvidia/nemotron-3-super-120b-a12b","name":"Nemotron 3 Super 120B A12B","description":"Nemotron middle tier for collaborative agents and high-volume reasoning workloads","family":"nemotron","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-06","release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144}},"nvidia/nemotron-3.5-lightning":{"id":"nvidia/nemotron-3.5-lightning","name":"Nemotron 3.5 Lightning 30B A3B","description":"Fast NVIDIA Nemotron MoE for reliable agentic tasks across enterprise workloads","family":"nemotron","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-08-11","last_updated":"2026-08-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"benchmarks":[{"name":"MMLU-Pro","score":81.94,"metric":"score","harness":"NeMo Gym / NeMo Evaluator SDK","variant":"BF16; reasoning","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16"},{"name":"AA-Omniscience","score":17.5,"metric":"score","harness":"NeMo Gym / NeMo Evaluator SDK","variant":"BF16; reasoning","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16"},{"name":"SciCode","score":32.6,"metric":"score","harness":"NeMo Gym / NeMo Evaluator SDK","variant":"BF16; reasoning","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16"},{"name":"PinchBench","score":85.37,"metric":"score","harness":"NeMo Gym / NeMo Evaluator SDK","variant":"BF16; reasoning","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16"},{"name":"BrowseComp","score":36.97,"metric":"score","harness":"NeMo Gym / NeMo Evaluator SDK","variant":"BF16; reasoning","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16"},{"name":"AA-LCR","score":52,"metric":"score","harness":"NeMo Gym / NeMo Evaluator SDK","variant":"BF16; reasoning","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16"},{"name":"GPQA Diamond","score":75.44,"metric":"accuracy","variant":"BF16; reasoning; no tools","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16"},{"name":"Humanity's Last Exam","score":11.72,"metric":"accuracy","variant":"BF16; reasoning; no tools","dataset":"text-only","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16"},{"name":"SWE-Bench Verified","score":51.56,"metric":"accuracy","harness":"NeMo Evaluator","variant":"BF16; reasoning","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16"},{"name":"SWE-Bench Multilingual","score":39.33,"metric":"accuracy","harness":"NeMo Evaluator","variant":"BF16; reasoning","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16"},{"name":"Terminal-Bench","score":24.58,"metric":"accuracy","harness":"NeMo Evaluator","variant":"BF16; reasoning","version":"2.1","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16"},{"name":"tau3-bench","score":9.28,"metric":"score","variant":"BF16; reasoning; banking","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16"},{"name":"GDPval-AA","score":832,"metric":"Elo","variant":"BF16; reasoning","version":"2","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16"},{"name":"IFBench","score":71.88,"metric":"accuracy","variant":"BF16; reasoning; loose","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16"}]},"nvidia/llama-3.1-nemotron-70b-instruct":{"id":"nvidia/llama-3.1-nemotron-70b-instruct","name":"Llama 3.1 Nemotron 70B Instruct","description":"Nemotron model for efficient reasoning, coding, and specialized AI agents","family":"nemotron","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-04-15","last_updated":"2025-04-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192}},"nvidia/nemotron-3.5-super-vl-120b-a12b":{"id":"nvidia/nemotron-3.5-super-vl-120b-a12b","name":"Nemotron 3.5 Super VL 120B A12B","description":"NVIDIA vision-language reasoning model for text, image, and video understanding and agentic tool use","family":"nemotron","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-10-15","last_updated":"2026-10-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":8192}},"nvidia/llama-nemotron-rerank-vl-1b-v2":{"id":"nvidia/llama-nemotron-rerank-vl-1b-v2","type":"reranking","name":"Llama Nemotron Rerank VL 1B v2","description":"Reranking model for improving retrieval quality in search and recommendation systems","family":"nemotron","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-18","last_updated":"2025-12-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":10240,"output":4096}},"nvidia/nemotron-3-nano-omni-30b-a3b-reasoning":{"id":"nvidia/nemotron-3-nano-omni-30b-a3b-reasoning","name":"Nemotron 3 Nano Omni 30B A3B Reasoning","description":"Open Nemotron omni model combining reasoning with text, vision, and audio","family":"nemotron","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-04-28","last_updated":"2026-04-28","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":65536}},"nvidia/nemotron-3-ultra-550b-a55b":{"id":"nvidia/nemotron-3-ultra-550b-a55b","name":"Nemotron 3 Ultra 550B A55B","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","family":"nemotron","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-06-04","last_updated":"2026-06-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":128000},"benchmarks":[{"name":"SWE-Bench Verified","score":70.7,"metric":"resolved","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16","date":"2026-06-04"},{"name":"SWE-Bench Multilingual","score":67.7,"metric":"resolve rate","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16","date":"2026-06-04"},{"name":"Terminal-Bench","score":56.4,"metric":"success rate","version":"2.1","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16","date":"2026-06-04"},{"name":"GPQA","score":87,"metric":"accuracy","variant":"no tools","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16","date":"2026-06-04"},{"name":"Humanity's Last Exam","score":26.7,"metric":"accuracy","variant":"no tools","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16","date":"2026-06-04"},{"name":"Humanity's Last Exam","score":37.4,"metric":"accuracy","variant":"with tools","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16","date":"2026-06-04"},{"name":"LiveCodeBench","score":89,"metric":"pass@1","version":"v6","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16","date":"2026-06-04"},{"name":"MMLU-Pro","score":86.8,"metric":"accuracy","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16","date":"2026-06-04"},{"name":"BrowseComp","score":44.4,"metric":"accuracy","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16","date":"2026-06-04"},{"name":"IFBench","score":81.7,"metric":"accuracy","variant":"prompt loose","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16","date":"2026-06-04"},{"name":"GDPval","score":46.7,"metric":"wins or ties","source":"https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16","date":"2026-06-04"}]},"nvidia/nemotron-voicechat":{"id":"nvidia/nemotron-voicechat","name":"Nemotron VoiceChat","description":"Nemotron multimodal model for visual reasoning and agentic AI workflows","family":"nemotron","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text","audio"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192}},"nvidia/llama-3.1-nemotron-safety-guard-8b-v3":{"id":"nvidia/llama-3.1-nemotron-safety-guard-8b-v3","name":"Llama 3.1 Nemotron Safety Guard 8B v3","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"nemotron","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-10-28","last_updated":"2025-10-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096}},"nvidia/nemotron-3-nano-30b-a3b":{"id":"nvidia/nemotron-3-nano-30b-a3b","name":"Nemotron 3 Nano 30B A3B","description":"Small Nemotron 3 MoE for efficient coding, math, and long-context agents","family":"nemotron","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-06-25","release_date":"2025-12-15","last_updated":"2025-12-15","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144}},"nvidia/mistral-nemotron":{"id":"nvidia/mistral-nemotron","name":"Mistral Nemotron","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"nemotron","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-06-11","last_updated":"2025-06-12","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":8192}},"nvidia/nemotron-cascade-2-30b-a3b":{"id":"nvidia/nemotron-cascade-2-30b-a3b","name":"Nemotron Cascade 2 30B A3B","description":"Nemotron model for efficient reasoning, coding, and specialized AI agents","family":"nemotron","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-03-19","last_updated":"2026-04-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":32768}},"nvidia/nemotron-nano-12b-v2-vl":{"id":"nvidia/nemotron-nano-12b-v2-vl","name":"Nemotron Nano 12B v2 VL","description":"Nemotron multimodal model for visual reasoning and agentic AI workflows","family":"nemotron","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2025-10-28","last_updated":"2025-10-28","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":128000}},"nvidia/llama-3.3-nemotron-super-49b-v1.5":{"id":"nvidia/llama-3.3-nemotron-super-49b-v1.5","name":"Llama 3.3 Nemotron Super 49B v1.5","description":"Nemotron model for efficient reasoning, coding, and specialized AI agents","family":"nemotron","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2025-07-25","last_updated":"2025-07-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072}},"nvidia/llama-3.1-nemotron-ultra-253b":{"id":"nvidia/llama-3.1-nemotron-ultra-253b","name":"Llama 3.1 Nemotron Ultra 253B","description":"Flagship Nemotron model for high-throughput reasoning and complex agents","family":"nemotron","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2025-04-07","last_updated":"2025-04-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192}},"nvidia/llama-nemotron-embed-vl-1b-v2":{"id":"nvidia/llama-nemotron-embed-vl-1b-v2","type":"embedding","name":"Llama Nemotron Embed VL 1B v2","description":"Embedding model for semantic search, retrieval, clustering, and ranking pipelines","family":"nemotron","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-18","last_updated":"2025-12-18","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":10240,"output":2048}},"nvidia/nemotron-3.5-content-safety":{"id":"nvidia/nemotron-3.5-content-safety","name":"Nemotron 3.5 Content Safety","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"nemotron","attachment":true,"reasoning":true,"tool_call":false,"temperature":true,"release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192}},"nvidia/nemotron-3-content-safety":{"id":"nvidia/nemotron-3-content-safety","name":"Nemotron 3 Content Safety","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"nemotron","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-03-16","last_updated":"2026-03-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4096}},"moonshotai/kimi-k2.7-code-highspeed":{"id":"moonshotai/kimi-k2.7-code-highspeed","name":"Kimi K2.7 Code Highspeed","description":"Lower-latency Kimi Code variant for interactive edits and coding-agent loops","family":"kimi-k2","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/moonshotai/Kimi-K2.7-Code"}]},"moonshotai/kimi-k2-thinking-turbo":{"id":"moonshotai/kimi-k2-thinking-turbo","name":"Kimi K2 Thinking Turbo","description":"Kimi reasoning model for long-horizon research, planning, and tool use","family":"kimi-thinking","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-11-06","last_updated":"2025-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/moonshotai/Kimi-K2-Thinking"}]},"moonshotai/kimi-k3":{"id":"moonshotai/kimi-k3","name":"Kimi K3","description":"Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work","family":"kimi-k3","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"release_date":"2026-07-16","last_updated":"2026-07-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"benchmarks":[{"name":"DeepSWE","score":67.5,"metric":"resolve rate","harness":"Kimi Code","variant":"max effort","version":"1.1","source":"https://www.kimi.com/blog/kimi-k3","date":"2026-07-16"},{"name":"Terminal-Bench","score":88.3,"metric":"accuracy","harness":"Kimi Code","variant":"max effort","version":"2.1","source":"https://www.kimi.com/blog/kimi-k3","date":"2026-07-16"},{"name":"FrontierSWE","score":81.2,"metric":"dominance score","harness":"Kimi Code","variant":"max effort","source":"https://www.kimi.com/blog/kimi-k3","date":"2026-07-16"},{"name":"Program Bench","score":77.8,"metric":"score","harness":"Kimi Code","variant":"max effort","source":"https://www.kimi.com/blog/kimi-k3","date":"2026-07-16"},{"name":"SWE Marathon","score":42,"metric":"resolve rate","harness":"Claude Code","variant":"max effort","version":"1.1","source":"https://www.kimi.com/blog/kimi-k3","date":"2026-07-16"},{"name":"GDPval-AA","score":1668,"metric":"Elo","variant":"max effort","version":"v2","source":"https://www.kimi.com/blog/kimi-k3","date":"2026-07-16"},{"name":"AA-Briefcase","score":1548,"metric":"Elo","variant":"max effort","source":"https://www.kimi.com/blog/kimi-k3","date":"2026-07-16"},{"name":"AutomationBench","score":30.8,"metric":"success rate","variant":"max effort","dataset":"600-task public subset","source":"https://www.kimi.com/blog/kimi-k3","date":"2026-07-16"},{"name":"JobBench","score":52.9,"metric":"score","variant":"max effort","source":"https://www.kimi.com/blog/kimi-k3","date":"2026-07-16"},{"name":"SpreadsheetBench","score":34.8,"metric":"score","harness":"Claude Code","variant":"max effort","version":"2","source":"https://www.kimi.com/blog/kimi-k3","date":"2026-07-16"},{"name":"BrowseComp","score":91.2,"metric":"accuracy","variant":"max effort, context compaction","source":"https://www.kimi.com/blog/kimi-k3","date":"2026-07-16"},{"name":"CharXiv Reasoning","score":91.3,"metric":"accuracy","variant":"max effort, with tools","source":"https://www.kimi.com/blog/kimi-k3","date":"2026-07-16"},{"name":"ZeroBench","score":41,"metric":"pass@5","variant":"max effort, with tools","source":"https://www.kimi.com/blog/kimi-k3","date":"2026-07-16"}]},"moonshotai/kimi-k2.6":{"id":"moonshotai/kimi-k2.6","name":"Kimi K2.6","description":"Multimodal Kimi workhorse for agent loops, coding tasks, and visual context","family":"kimi-k2","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-01","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/moonshotai/Kimi-K2.6"}],"benchmarks":[{"name":"SWE-Bench Verified","score":80.2,"metric":"resolved","source":"https://huggingface.co/moonshotai/Kimi-K2.6"},{"name":"Artificial Analysis Coding Agent Index","score":50.5,"metric":"average pass@1","harness":"Claude Code","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"SWE-Atlas Codebase QnA","score":59.8,"metric":"pass@1","harness":"Claude Code","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"SWE-Bench Pro","score":27.3,"metric":"pass@1","harness":"Claude Code","dataset":"hard-aa","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"Terminal-Bench","score":64.3,"metric":"pass@1","harness":"Claude Code","version":"2.1","source":"https://artificialanalysis.ai/agents/coding-agents"}]},"moonshotai/kimi-k2":{"id":"moonshotai/kimi-k2","name":"Kimi K2 0711","description":"Kimi model for long-context chat, coding, and agentic reasoning","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-12-31","release_date":"2025-07-11","last_updated":"2025-07-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":98304}},"moonshotai/kimi-k2.5":{"id":"moonshotai/kimi-k2.5","name":"Kimi K2.5","description":"Earlier Kimi frontier model for long-context agents, coding, and multimodal work","family":"kimi-k2","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/moonshotai/Kimi-K2.5"}],"benchmarks":[{"name":"SWE-Bench Verified","score":70.8,"metric":"resolved","source":"https://www.swebench.com/"},{"name":"SWE-Atlas Codebase QnA","score":13.1,"metric":"score","harness":"Mini-SWE-Agent","source":"https://labs.scale.com/leaderboard/sweatlas-qna"},{"name":"SWE-Atlas Refactoring","score":20.95,"metric":"score","harness":"Mini-SWE-Agent","source":"https://labs.scale.com/leaderboard/sweatlas-refactoring"},{"name":"SWE-Atlas Test Writing","score":25.77,"metric":"score","harness":"Mini-SWE-Agent","source":"https://labs.scale.com/leaderboard/sweatlas-tw"}]},"moonshotai/kimi-k2.8-preview":{"id":"moonshotai/kimi-k2.8-preview","name":"Kimi K2.8 Preview","description":"Kimi coding model with more efficient thinking and up to 1M context, available through Kimi Code","family":"kimi-k2","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"release_date":"2026-09-11","last_updated":"2026-09-11","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576}},"moonshotai/kimi-k3.1-pro":{"id":"moonshotai/kimi-k3.1-pro","name":"Kimi K3.1 Pro","description":"Multimodal Kimi model with 1M context and always-on reasoning for coding, 3D creation, financial research, and agentic work","family":"kimi-k3","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"release_date":"2026-10-23","last_updated":"2026-10-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072}},"moonshotai/kimi-k2-0905":{"id":"moonshotai/kimi-k2-0905","name":"Kimi K2 Instruct 0905","description":"September 2025 Kimi K2 instruction update with a 256K context and improved agentic coding","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-09-05","last_updated":"2025-09-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":98304},"license":"Modified MIT","weights":[{"label":"Hugging Face","url":"https://huggingface.co/moonshotai/Kimi-K2-Instruct-0905"}]},"moonshotai/kimi-k2.7-code":{"id":"moonshotai/kimi-k2.7-code","name":"Kimi K2.7 Code","description":"Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking","family":"kimi-k2","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2025-01","release_date":"2026-06-12","last_updated":"2026-06-12","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/moonshotai/Kimi-K2.7-Code"}],"benchmarks":[{"name":"Kimi Code Bench","score":62,"harness":"Kimi Code CLI","version":"v2","source":"https://huggingface.co/moonshotai/Kimi-K2.7-Code","date":"2026-06-12"},{"name":"Program Bench","score":53.6,"harness":"Kimi Code CLI","source":"https://huggingface.co/moonshotai/Kimi-K2.7-Code","date":"2026-06-12"},{"name":"MLS Bench Lite","score":35.1,"harness":"Kimi Code CLI","source":"https://huggingface.co/moonshotai/Kimi-K2.7-Code","date":"2026-06-12"},{"name":"MCP Atlas","score":76,"metric":"success rate","harness":"Kimi Code CLI","source":"https://huggingface.co/moonshotai/Kimi-K2.7-Code","date":"2026-06-12"},{"name":"MCP Mark Verified","score":81.1,"metric":"success rate","harness":"Kimi Code CLI","source":"https://huggingface.co/moonshotai/Kimi-K2.7-Code","date":"2026-06-12"},{"name":"Kimi Claw 24/7 Bench","score":46.9,"harness":"Kimi Code CLI","source":"https://huggingface.co/moonshotai/Kimi-K2.7-Code","date":"2026-06-12"}]},"moonshotai/kimi-k2-instruct":{"id":"moonshotai/kimi-k2-instruct","name":"Kimi K2 Instruct","description":"Non-thinking Kimi K2 MoE instruction model for coding and tool-using agents","family":"kimi-k2","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-07-11","last_updated":"2025-07-11","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"license":"Modified MIT","weights":[{"label":"Hugging Face","url":"https://huggingface.co/moonshotai/Kimi-K2-Instruct"}]},"moonshotai/kimi-k2-thinking":{"id":"moonshotai/kimi-k2-thinking","name":"Kimi K2 Thinking","description":"Thinking Kimi model for slower research passes, planning, and hard technical questions","family":"kimi-thinking","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2024-08","release_date":"2025-11-06","last_updated":"2025-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/moonshotai/Kimi-K2-Thinking"}],"benchmarks":[{"name":"SWE-Bench Verified","score":71.3,"metric":"resolved","source":"https://huggingface.co/moonshotai/Kimi-K2-Thinking"}]},"vispark/vision-medium":{"id":"vispark/vision-medium","name":"Vision Medium","description":"Balanced multimodal model pairing a 1M-token context window with deeper reasoning for analysis, content creation, and tool use across text, image, audio, video, and PDF inputs.","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2024-05-15","last_updated":"2026-09","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536}},"vispark/vision-large":{"id":"vispark/vision-large","name":"Vision Large","description":"Most capable Vision model for complex reasoning, detailed media analysis, and structured output over a 1M-token context window.","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2024-05-15","last_updated":"2026-09","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536}},"vispark/vision-small":{"id":"vispark/vision-small","name":"Vision Small","description":"Fast, low-cost multimodal model for understanding text, images, audio, video, and PDFs, with tool calling and a 1M-token context window.","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2024-05-15","last_updated":"2026-09","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536}},"canopylabs/orpheus-arabic-saudi":{"id":"canopylabs/orpheus-arabic-saudi","type":"speech","name":"Orpheus Arabic Saudi","description":"Expressive Saudi Arabic text-to-speech model with bracketed vocal directions","family":"canopylabs","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-16","last_updated":"2025-12-16","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":4000,"output":50000}},"canopylabs/orpheus-v1-english":{"id":"canopylabs/orpheus-v1-english","type":"speech","name":"Orpheus V1 English","description":"Expressive English text-to-speech model with bracketed vocal directions","family":"canopylabs","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-12-19","last_updated":"2025-12-19","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":4000,"output":50000}},"openai/gpt-4o-mini":{"id":"openai/gpt-4o-mini","name":"GPT-4o mini","description":"Small omni GPT for cheap multimodal assistance and production-scale traffic","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-07-18","last_updated":"2024-07-18","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"benchmarks":[{"name":"Aider Polyglot","score":3.6,"metric":"percent correct","source":"https://aider.chat/docs/leaderboards/","date":"2024-12-21"},{"name":"SciCode","score":22.9,"metric":"percent correct","source":"https://openrouter.ai/openai/gpt-4o-mini/benchmarks","date":"2026-03-11"}]},"openai/whisper-large-v3":{"id":"openai/whisper-large-v3","type":"transcription","name":"Whisper 3 Large","description":"Open Whisper checkpoint for robust multilingual transcription and captioning","family":"whisper","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2023-11-06","last_updated":"2023-11-06","modalities":{"input":["audio"],"output":["text"]},"open_weights":true,"limit":{"context":448,"output":4096}},"openai/o3-pro":{"id":"openai/o3-pro","name":"o3-pro","description":"High-effort o3 tier for difficult technical reasoning and careful answers","family":"o-pro","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-06-10","last_updated":"2025-06-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"benchmarks":[{"name":"Aider Polyglot","score":84.9,"metric":"percent correct","source":"https://aider.chat/docs/leaderboards/","date":"2025-06-28"}]},"openai/gpt-5.6-cyber":{"id":"openai/gpt-5.6-cyber","name":"GPT-5.6 Cyber","description":"Cybersecurity model for authorized vulnerability research and security testing","family":"gpt","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-08-07","last_updated":"2026-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000}},"openai/gpt-oss-20b":{"id":"openai/gpt-oss-20b","name":"GPT OSS 20B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/openai/gpt-oss-20b"}]},"openai/gpt-image-2.5-sunburst":{"id":"openai/gpt-image-2.5-sunburst","type":"image","name":"GPT Image 2.5 Sunburst","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-08","last_updated":"2026-09-08","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0}},"openai/gpt-6-sol":{"id":"openai/gpt-6-sol","name":"GPT-6 Sol","description":"OpenAI model for complex coding and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2026-04-20","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"benchmarks":[{"name":"DeepSWE","score":68.8,"metric":"score","variant":"max effort","version":"1.1","source":"https://openai.com/index/introducing-gpt-6-sol-and-luna/","date":"2026-09-22"},{"name":"AutomationBench","score":33.2,"metric":"score","variant":"xhigh effort","version":"1.0.6","source":"https://openai.com/index/introducing-gpt-6-sol-and-luna/","date":"2026-09-22"},{"name":"Agents' Last Exam","score":56.4,"metric":"score","variant":"max effort","version":"1","source":"https://openai.com/index/introducing-gpt-6-sol-and-luna/","date":"2026-09-22"},{"name":"OSWorld","score":60.5,"metric":"partial reward","variant":"xhigh effort","dataset":"offline v2026.08.08","version":"2.0","source":"https://openai.com/index/introducing-gpt-6-sol-and-luna/","date":"2026-09-22"}]},"openai/gpt-5-nano":{"id":"openai/gpt-5-nano","name":"GPT-5 Nano","description":"Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs","family":"gpt-nano","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000}},"openai/o3":{"id":"openai/o3","name":"o3","description":"Deliberate o-series reasoner for hard math, coding, and multi-step analysis","family":"o","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"benchmarks":[{"name":"Aider Polyglot","score":81.3,"metric":"percent correct","source":"https://aider.chat/docs/leaderboards/","date":"2025-06-25"}]},"openai/o1":{"id":"openai/o1","name":"o1","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2023-09","release_date":"2024-12-05","last_updated":"2024-12-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"benchmarks":[{"name":"Aider Polyglot","score":61.7,"metric":"percent correct","source":"https://aider.chat/docs/leaderboards/","date":"2024-12-21"}]},"openai/gpt-5-chat-latest":{"id":"openai/gpt-5-chat-latest","name":"GPT-5 Chat (latest)","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"tool_call":false,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000}},"openai/gpt-5.2-chat-latest":{"id":"openai/gpt-5.2-chat-latest","name":"GPT-5.2 Chat","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384}},"openai/gpt-5.1":{"id":"openai/gpt-5.1","name":"GPT-5.1","description":"Sharper GPT-5 generation for coding, product work, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000}},"openai/gpt-5.3-codex-spark":{"id":"openai/gpt-5.3-codex-spark","name":"GPT-5.3 Codex Spark","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex-spark","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"input":100000,"output":32000}},"openai/gpt-oss-safeguard-20b":{"id":"openai/gpt-oss-safeguard-20b","name":"GPT OSS Safeguard 20B","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"gpt-oss","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2025-10-29","last_updated":"2025-10-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":131072},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/openai/gpt-oss-safeguard-20b"}]},"openai/gpt-5.3-codex":{"id":"openai/gpt-5.3-codex","name":"GPT-5.3 Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-05","last_updated":"2026-02-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"benchmarks":[{"name":"SWE-Atlas Codebase QnA","score":32.6,"metric":"score","harness":"Codex","source":"https://labs.scale.com/leaderboard/sweatlas-qna"},{"name":"SWE-Atlas Refactoring","score":42.38,"metric":"score","harness":"Codex","source":"https://labs.scale.com/leaderboard/sweatlas-refactoring"},{"name":"SWE-Atlas Test Writing","score":38.98,"metric":"score","harness":"Codex","source":"https://labs.scale.com/leaderboard/sweatlas-tw"}]},"openai/gpt-5.4-image-2":{"id":"openai/gpt-5.4-image-2","type":"image","name":"GPT-5.4 Image 2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt","attachment":true,"reasoning":true,"tool_call":false,"temperature":false,"release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["image","text","pdf"],"output":["image","text"]},"open_weights":false,"limit":{"context":272000,"output":128000}},"openai/gpt-4o-2024-08-06":{"id":"openai/gpt-4o-2024-08-06","name":"GPT-4o (2024-08-06)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-08-06","last_updated":"2024-08-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"benchmarks":[{"name":"Aider Polyglot","score":23.1,"metric":"percent correct","source":"https://aider.chat/docs/leaderboards/","date":"2024-12-30"},{"name":"Artificial Analysis Coding Index","score":16.6,"metric":"index","source":"https://openrouter.ai/openai/gpt-4o-2024-08-06/benchmarks","date":"2026-03-11"},{"name":"SciCode","score":33.1,"metric":"percent correct","source":"https://openrouter.ai/openai/gpt-4o-2024-08-06/benchmarks","date":"2026-03-11"},{"name":"Terminal-Bench Hard","score":8.3,"metric":"success rate","source":"https://openrouter.ai/openai/gpt-4o-2024-08-06/benchmarks","date":"2026-03-11"}]},"openai/gpt-5.4-mini":{"id":"openai/gpt-5.4-mini","name":"GPT-5.4 mini","description":"Strong small GPT for coding subagents, quick tool use, and high-volume work","family":"gpt-mini","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"benchmarks":[{"name":"SWE-Bench Pro","score":54.4,"metric":"resolve rate","variant":"reasoning effort xhigh","source":"https://openai.com/index/introducing-gpt-5-4-mini-and-nano/","date":"2026-03-17"},{"name":"Terminal-Bench","score":60,"metric":"accuracy","variant":"reasoning effort xhigh","version":"2.0","source":"https://openai.com/index/introducing-gpt-5-4-mini-and-nano/","date":"2026-03-17"},{"name":"MCP Atlas","score":57.7,"metric":"score","variant":"reasoning effort xhigh","source":"https://openai.com/index/introducing-gpt-5-4-mini-and-nano/","date":"2026-03-17"},{"name":"Toolathlon","score":42.9,"metric":"score","variant":"reasoning effort xhigh","source":"https://openai.com/index/introducing-gpt-5-4-mini-and-nano/","date":"2026-03-17"},{"name":"τ²-Bench Telecom","score":93.4,"metric":"accuracy","variant":"reasoning effort xhigh","source":"https://openai.com/index/introducing-gpt-5-4-mini-and-nano/","date":"2026-03-17"},{"name":"GPQA Diamond","score":88,"metric":"accuracy","variant":"reasoning effort xhigh","source":"https://openai.com/index/introducing-gpt-5-4-mini-and-nano/","date":"2026-03-17"},{"name":"Humanity's Last Exam","score":41.5,"metric":"accuracy","variant":"with tools","source":"https://openai.com/index/introducing-gpt-5-4-mini-and-nano/","date":"2026-03-17"},{"name":"Humanity's Last Exam","score":28.2,"metric":"accuracy","variant":"without tools","source":"https://openai.com/index/introducing-gpt-5-4-mini-and-nano/","date":"2026-03-17"},{"name":"OSWorld-Verified","score":72.1,"metric":"success rate","variant":"reasoning effort xhigh","source":"https://openai.com/index/introducing-gpt-5-4-mini-and-nano/","date":"2026-03-17"},{"name":"MMMU Pro","score":78,"metric":"accuracy","variant":"with Python","source":"https://openai.com/index/introducing-gpt-5-4-mini-and-nano/","date":"2026-03-17"},{"name":"MMMU Pro","score":76.6,"metric":"accuracy","variant":"without tools","source":"https://openai.com/index/introducing-gpt-5-4-mini-and-nano/","date":"2026-03-17"},{"name":"OmniDocBench","score":0.1263,"metric":"overall edit distance","variant":"reasoning effort none","version":"1.5","source":"https://openai.com/index/introducing-gpt-5-4-mini-and-nano/","date":"2026-03-17"},{"name":"OpenAI MRCR","score":47.7,"metric":"accuracy","variant":"8-needle, 64K-128K","version":"v2","source":"https://openai.com/index/introducing-gpt-5-4-mini-and-nano/","date":"2026-03-17"},{"name":"OpenAI MRCR","score":33.6,"metric":"accuracy","variant":"8-needle, 128K-256K","version":"v2","source":"https://openai.com/index/introducing-gpt-5-4-mini-and-nano/","date":"2026-03-17"},{"name":"Graphwalks","score":76.3,"metric":"accuracy","variant":"BFS, 0-128K","source":"https://openai.com/index/introducing-gpt-5-4-mini-and-nano/","date":"2026-03-17"},{"name":"Graphwalks","score":71.5,"metric":"accuracy","variant":"parents, 0-128K","source":"https://openai.com/index/introducing-gpt-5-4-mini-and-nano/","date":"2026-03-17"}]},"openai/gpt-3.5-turbo":{"id":"openai/gpt-3.5-turbo","name":"GPT-3.5-turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2021-09-01","release_date":"2023-03-01","last_updated":"2023-11-06","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16385,"output":4096},"benchmarks":[{"name":"Artificial Analysis Coding Index","score":10.7,"metric":"index","source":"https://openrouter.ai/openai/gpt-3.5-turbo/benchmarks","date":"2026-03-11"}]},"openai/gpt-4":{"id":"openai/gpt-4","name":"GPT-4","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-11","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":8192,"output":8192},"benchmarks":[{"name":"Artificial Analysis Coding Index","score":13.1,"metric":"index","source":"https://openrouter.ai/openai/gpt-4/benchmarks","date":"2026-03-11"}]},"openai/gpt-5-pro":{"id":"openai/gpt-5-pro","name":"GPT-5 Pro","description":"Higher-accuracy GPT-5 tier for tough analysis, coding reviews, and planning","family":"gpt-pro","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-10-06","last_updated":"2025-10-06","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":272000}},"openai/gpt-5.6-terra":{"id":"openai/gpt-5.6-terra","name":"GPT-5.6 Terra","description":"Balanced GPT-5.6 model for capable, cost-efficient everyday work","family":"gpt-terra","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"benchmarks":[{"name":"SWE-Bench Pro","score":63.4,"metric":"resolve rate","source":"https://openai.com/index/gpt-5-6/","date":"2026-07-09"},{"name":"Terminal-Bench","score":87.4,"metric":"success rate","version":"2.1","source":"https://openai.com/index/gpt-5-6/","date":"2026-07-09"},{"name":"DeepSWE","score":69.6,"metric":"resolve rate","version":"1.1","source":"https://openai.com/index/gpt-5-6/","date":"2026-07-09"},{"name":"GPQA Diamond","score":92.9,"metric":"accuracy","source":"https://openai.com/index/gpt-5-6/","date":"2026-07-09"},{"name":"FrontierMath","score":84.9,"metric":"accuracy","dataset":"Tier 1-3","version":"v2","source":"https://openai.com/index/gpt-5-6/","date":"2026-07-09"},{"name":"BrowseComp","score":87.5,"metric":"accuracy","source":"https://openai.com/index/gpt-5-6/","date":"2026-07-09"},{"name":"OSWorld","score":50.2,"metric":"success rate","version":"2.0","source":"https://openai.com/index/gpt-5-6/","date":"2026-07-09"},{"name":"MMMU Pro","score":80.7,"metric":"accuracy","variant":"no tools","source":"https://openai.com/index/gpt-5-6/","date":"2026-07-09"},{"name":"Agents' Last Exam","score":50.4,"source":"https://openai.com/index/gpt-5-6/","date":"2026-07-09"},{"name":"Toolathlon","score":53.1,"metric":"success rate","source":"https://openai.com/index/gpt-5-6/","date":"2026-07-09"},{"name":"Artificial Analysis Intelligence Index","score":55,"metric":"index score","variant":"max","version":"4.1","source":"https://artificialanalysis.ai/articles/gpt-5-6-has-landed","date":"2026-07-09"},{"name":"Artificial Analysis Coding Agent Index","score":77.4,"metric":"index score","harness":"Codex","variant":"max","version":"1.1","source":"https://artificialanalysis.ai/articles/gpt-5-6-has-landed","date":"2026-07-09"}]},"openai/gpt-5.4-nano":{"id":"openai/gpt-5.4-nano","name":"GPT-5.4 nano","description":"Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation","family":"gpt-nano","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-17","last_updated":"2026-03-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"benchmarks":[{"name":"SWE-Bench Pro","score":52.4,"metric":"resolve rate","variant":"reasoning effort xhigh","source":"https://openai.com/index/introducing-gpt-5-4-mini-and-nano/","date":"2026-03-17"},{"name":"Terminal-Bench","score":46.3,"metric":"accuracy","variant":"reasoning effort xhigh","version":"2.0","source":"https://openai.com/index/introducing-gpt-5-4-mini-and-nano/","date":"2026-03-17"},{"name":"MCP Atlas","score":56.1,"metric":"score","variant":"reasoning effort xhigh","source":"https://openai.com/index/introducing-gpt-5-4-mini-and-nano/","date":"2026-03-17"},{"name":"Toolathlon","score":35.5,"metric":"score","variant":"reasoning effort xhigh","source":"https://openai.com/index/introducing-gpt-5-4-mini-and-nano/","date":"2026-03-17"},{"name":"τ²-Bench Telecom","score":92.5,"metric":"accuracy","variant":"reasoning effort xhigh","source":"https://openai.com/index/introducing-gpt-5-4-mini-and-nano/","date":"2026-03-17"},{"name":"GPQA Diamond","score":82.8,"metric":"accuracy","variant":"reasoning effort xhigh","source":"https://openai.com/index/introducing-gpt-5-4-mini-and-nano/","date":"2026-03-17"},{"name":"Humanity's Last Exam","score":37.7,"metric":"accuracy","variant":"with tools","source":"https://openai.com/index/introducing-gpt-5-4-mini-and-nano/","date":"2026-03-17"},{"name":"Humanity's Last Exam","score":24.3,"metric":"accuracy","variant":"without tools","source":"https://openai.com/index/introducing-gpt-5-4-mini-and-nano/","date":"2026-03-17"},{"name":"OSWorld-Verified","score":39,"metric":"success rate","variant":"reasoning effort xhigh","source":"https://openai.com/index/introducing-gpt-5-4-mini-and-nano/","date":"2026-03-17"},{"name":"MMMU Pro","score":69.5,"metric":"accuracy","variant":"with Python","source":"https://openai.com/index/introducing-gpt-5-4-mini-and-nano/","date":"2026-03-17"},{"name":"MMMU Pro","score":66.1,"metric":"accuracy","variant":"without tools","source":"https://openai.com/index/introducing-gpt-5-4-mini-and-nano/","date":"2026-03-17"},{"name":"OmniDocBench","score":0.2419,"metric":"overall edit distance","variant":"reasoning effort none","version":"1.5","source":"https://openai.com/index/introducing-gpt-5-4-mini-and-nano/","date":"2026-03-17"},{"name":"OpenAI MRCR","score":44.2,"metric":"accuracy","variant":"8-needle, 64K-128K","version":"v2","source":"https://openai.com/index/introducing-gpt-5-4-mini-and-nano/","date":"2026-03-17"},{"name":"OpenAI MRCR","score":33.1,"metric":"accuracy","variant":"8-needle, 128K-256K","version":"v2","source":"https://openai.com/index/introducing-gpt-5-4-mini-and-nano/","date":"2026-03-17"},{"name":"Graphwalks","score":73.4,"metric":"accuracy","variant":"BFS, 0-128K","source":"https://openai.com/index/introducing-gpt-5-4-mini-and-nano/","date":"2026-03-17"},{"name":"Graphwalks","score":50.8,"metric":"accuracy","variant":"parents, 0-128K","source":"https://openai.com/index/introducing-gpt-5-4-mini-and-nano/","date":"2026-03-17"}]},"openai/gpt-4o":{"id":"openai/gpt-4o","name":"GPT-4o","description":"Omni-era GPT for multimodal chat, practical coding, and general assistants","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-08-06","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"benchmarks":[{"name":"Aider Polyglot","score":23.1,"metric":"percent correct","source":"https://aider.chat/docs/leaderboards/","date":"2024-12-30"}]},"openai/gpt-5":{"id":"openai/gpt-5","name":"GPT-5","description":"Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"benchmarks":[{"name":"Aider Polyglot","score":88,"metric":"percent correct","source":"https://aider.chat/docs/leaderboards/","date":"2025-08-23"},{"name":"SWE-Bench Pro","score":41.78,"metric":"resolve rate","dataset":"public","source":"https://labs.scale.com/leaderboard/swe_bench_pro_public"}]},"openai/gpt-4.1-nano":{"id":"openai/gpt-4.1-nano","name":"GPT-4.1 nano","description":"Tiny GPT-4.1 option for classification, routing, and very high-volume tasks","family":"gpt-nano","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"benchmarks":[{"name":"Aider Polyglot","score":8.9,"metric":"percent correct","source":"https://aider.chat/docs/leaderboards/","date":"2025-04-14"}]},"openai/gpt-4o-2024-05-13":{"id":"openai/gpt-4o-2024-05-13","name":"GPT-4o (2024-05-13)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-05-13","last_updated":"2024-05-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"benchmarks":[{"name":"Artificial Analysis Coding Index","score":24.2,"metric":"index","source":"https://openrouter.ai/openai/gpt-4o-2024-05-13/benchmarks","date":"2026-03-11"},{"name":"SciCode","score":30.9,"metric":"percent correct","source":"https://openrouter.ai/openai/gpt-4o-2024-05-13/benchmarks","date":"2026-03-11"}]},"openai/gpt-5.1-codex-max":{"id":"openai/gpt-5.1-codex-max","name":"GPT-5.1 Codex Max","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000}},"openai/gpt-4-turbo":{"id":"openai/gpt-4-turbo","name":"GPT-4 Turbo","description":"Compact GPT model for low-latency assistance and high-volume workloads","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-12","release_date":"2023-11-06","last_updated":"2024-04-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"benchmarks":[{"name":"Artificial Analysis Coding Index","score":21.5,"metric":"index","source":"https://openrouter.ai/openai/gpt-4-turbo/benchmarks","date":"2026-03-11"},{"name":"SciCode","score":31.9,"metric":"percent correct","source":"https://openrouter.ai/openai/gpt-4-turbo/benchmarks","date":"2026-03-11"}]},"openai/gpt-5.1-chat-latest":{"id":"openai/gpt-5.1-chat-latest","name":"GPT-5.1 Chat","description":"Chat-tuned GPT-5.1 for polished assistants, writing, and product conversations","family":"gpt","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384}},"openai/gpt-oss-safeguard-120b":{"id":"openai/gpt-oss-safeguard-120b","name":"GPT OSS Safeguard 120B","description":"Safety model for policy screening, moderation, and risk-aware routing workflows","family":"gpt-oss","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2025-10-29","last_updated":"2025-10-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/openai/gpt-oss-safeguard-120b"}]},"openai/gpt-5.5":{"id":"openai/gpt-5.5","name":"GPT-5.5","description":"Default frontier GPT for coding, computer use, research, and knowledge work","family":"gpt","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"benchmarks":[{"name":"SWE-Bench Pro","score":58.6,"metric":"resolve rate","source":"https://www.anthropic.com/news/claude-opus-4-8","date":"2026-05-28"},{"name":"Terminal-Bench","score":78.2,"metric":"success rate","harness":"Terminus-2","version":"2.1","source":"https://www.anthropic.com/news/claude-opus-4-8","date":"2026-05-28"},{"name":"SWE-Atlas Codebase QnA","score":45.43,"metric":"score","harness":"Codex","source":"https://labs.scale.com/leaderboard/sweatlas-qna"},{"name":"SWE-Atlas Refactoring","score":44.79,"metric":"score","harness":"Codex","source":"https://labs.scale.com/leaderboard/sweatlas-refactoring"},{"name":"SWE-Atlas Test Writing","score":42.59,"metric":"score","harness":"Codex","source":"https://labs.scale.com/leaderboard/sweatlas-tw"},{"name":"Artificial Analysis Coding Agent Index","score":65.3,"metric":"average pass@1","harness":"Codex","variant":"xhigh","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"SWE-Atlas Codebase QnA","score":80.8,"metric":"pass@1","harness":"Codex","variant":"xhigh","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"SWE-Bench Pro","score":30.9,"metric":"pass@1","harness":"Codex","variant":"xhigh","dataset":"hard-aa","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"Terminal-Bench","score":84.1,"metric":"pass@1","harness":"Codex","variant":"xhigh","version":"2.1","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"Artificial Analysis Coding Agent Index","score":60.4,"metric":"average pass@1","harness":"Codex","variant":"medium","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"SWE-Atlas Codebase QnA","score":79.1,"metric":"pass@1","harness":"Codex","variant":"medium","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"SWE-Bench Pro","score":26.2,"metric":"pass@1","harness":"Codex","variant":"medium","dataset":"hard-aa","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"Terminal-Bench","score":75.8,"metric":"pass@1","harness":"Codex","variant":"medium","version":"2.1","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"Artificial Analysis Coding Agent Index","score":57.8,"metric":"average pass@1","harness":"Cursor CLI","variant":"medium","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"SWE-Atlas Codebase QnA","score":75,"metric":"pass@1","harness":"Cursor CLI","variant":"medium","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"SWE-Bench Pro","score":24.9,"metric":"pass@1","harness":"Cursor CLI","variant":"medium","dataset":"hard-aa","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"Terminal-Bench","score":73.4,"metric":"pass@1","harness":"Cursor CLI","variant":"medium","version":"2.1","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"Terminal-Bench","score":82.7,"metric":"success rate","version":"2.0","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"GPQA Diamond","score":93.6,"metric":"accuracy","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"Humanity's Last Exam","score":41.4,"metric":"accuracy","variant":"no tools","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"Humanity's Last Exam","score":52.2,"metric":"accuracy","variant":"with tools","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"OSWorld-Verified","score":78.7,"metric":"success rate","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"BrowseComp","score":84.4,"metric":"accuracy","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"MMMU Pro","score":81.2,"metric":"accuracy","variant":"no tools","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"ARC-AGI-2","score":85,"metric":"accuracy","variant":"Verified","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"FrontierMath","score":51.7,"metric":"accuracy","dataset":"Tier 1-3","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"FrontierMath","score":35.4,"metric":"accuracy","dataset":"Tier 4","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"GDPval","score":84.9,"metric":"wins or ties","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"MCP Atlas","score":75.3,"metric":"success rate","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"Toolathlon","score":55.6,"metric":"success rate","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"τ²-Bench Telecom","score":98,"metric":"success rate","variant":"original prompts","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"}]},"openai/gpt-5.4-pro":{"id":"openai/gpt-5.4-pro","name":"GPT-5.4 Pro","description":"More exact GPT-5.4 tier for demanding professional reasoning and agent tasks","family":"gpt-pro","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"benchmarks":[{"name":"GPQA Diamond","score":94.4,"metric":"accuracy","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"Humanity's Last Exam","score":42.7,"metric":"accuracy","variant":"no tools","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"Humanity's Last Exam","score":58.7,"metric":"accuracy","variant":"with tools","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"BrowseComp","score":89.3,"metric":"accuracy","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"GDPval","score":82,"metric":"wins or ties","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"FrontierMath","score":50,"metric":"accuracy","dataset":"Tier 1-3","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"FrontierMath","score":38,"metric":"accuracy","dataset":"Tier 4","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"ARC-AGI-1","score":94.5,"metric":"accuracy","variant":"Verified","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"ARC-AGI-2","score":83.3,"metric":"accuracy","variant":"Verified","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"FinanceAgent","score":61.5,"metric":"accuracy","version":"1.1","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"GeneBench","score":25.6,"metric":"accuracy","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"}]},"openai/gpt-5.1-codex":{"id":"openai/gpt-5.1-codex","name":"GPT-5.1 Codex","description":"Codex GPT for repository edits, code review, and practical software agents","family":"gpt-codex","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000}},"openai/gpt-image-2.5-flare":{"id":"openai/gpt-image-2.5-flare","type":"image","name":"GPT Image 2.5 Flare","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-08","last_updated":"2026-09-08","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0}},"openai/gpt-realtime-whisper":{"id":"openai/gpt-realtime-whisper","type":"transcription","name":"GPT Realtime Whisper","description":"Streaming speech-to-text model for low-latency transcript deltas from live audio","family":"whisper","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-05-07","last_updated":"2026-05-07","modalities":{"input":["audio"],"output":["text"]},"open_weights":false,"limit":{"context":16000,"output":2000}},"openai/gpt-6-astra-fast":{"id":"openai/gpt-6-astra-fast","name":"GPT-6 Astra (Fast)","description":"Fast variant of GPT-6 Astra for low-latency assistance and high-volume workloads.","family":"gpt-astra","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000}},"openai/gpt-image-2":{"id":"openai/gpt-image-2","type":"image","name":"GPT-Image-2","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-21","last_updated":"2026-04-21","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0}},"openai/gpt-5-image":{"id":"openai/gpt-5-image","type":"image","name":"GPT-5 Image","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt","attachment":true,"reasoning":true,"tool_call":false,"temperature":true,"knowledge":"2024-10-01","release_date":"2025-10-14","last_updated":"2025-10-14","modalities":{"input":["image","text","pdf"],"output":["image","text"]},"open_weights":false,"limit":{"context":400000,"output":128000}},"openai/gpt-5.3-chat-latest":{"id":"openai/gpt-5.3-chat-latest","name":"GPT-5.3 Chat (latest)","description":"Chat-tuned GPT model for conversational assistance, writing, and tool workflows","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-03","last_updated":"2026-03-03","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384}},"openai/whisper-large-v3-turbo":{"id":"openai/whisper-large-v3-turbo","type":"transcription","name":"Whisper Large v3 Turbo","description":"Speech transcription model for accurate audio-to-text and captioning workflows","family":"whisper","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2024-10-01","last_updated":"2024-10-01","modalities":{"input":["audio"],"output":["text"]},"open_weights":true,"limit":{"context":448,"output":448}},"openai/gpt-6.1-terra":{"id":"openai/gpt-6.1-terra","name":"GPT-6.1 Terra","description":"Fast, efficient model for visual reasoning, precise computer use, and clear, natural writing.","family":"gpt-terra","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2026-06-01","release_date":"2026-10-26","last_updated":"2026-10-26","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000}},"openai/o1-pro":{"id":"openai/o1-pro","name":"o1-pro","description":"O-series reasoning model for hard analysis, math, coding, and planning","family":"o-pro","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2023-09","release_date":"2025-03-19","last_updated":"2025-03-19","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000}},"openai/gpt-5.5-pro":{"id":"openai/gpt-5.5-pro","name":"GPT-5.5 Pro","description":"Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding","family":"gpt-pro","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2025-12-01","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"benchmarks":[{"name":"BrowseComp","score":90.1,"metric":"accuracy","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"Humanity's Last Exam","score":43.1,"metric":"accuracy","variant":"no tools","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"Humanity's Last Exam","score":57.2,"metric":"accuracy","variant":"with tools","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"FrontierMath","score":52.4,"metric":"accuracy","dataset":"Tier 1-3","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"FrontierMath","score":39.6,"metric":"accuracy","dataset":"Tier 4","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"GDPval","score":82.3,"metric":"wins or ties","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"GeneBench","score":33.2,"metric":"accuracy","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"}]},"openai/gpt-5.2-codex":{"id":"openai/gpt-5.2-codex","name":"GPT-5.2 Codex","description":"Code-specialist GPT for repository edits, reviews, and long-running software agents","family":"gpt-codex","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"benchmarks":[{"name":"SWE-Bench Pro","score":41.04,"metric":"resolve rate","dataset":"public","source":"https://labs.scale.com/leaderboard/swe_bench_pro_public"}]},"openai/o4-mini-deep-research":{"id":"openai/o4-mini-deep-research","name":"o4-mini-deep-research","description":"Research model for long-horizon investigation, synthesis, and analytical reports","family":"o-mini","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-06-26","last_updated":"2025-06-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000}},"openai/gpt-4.1-mini":{"id":"openai/gpt-4.1-mini","name":"GPT-4.1 mini","description":"Affordable GPT-4.1 lane for fast coding help and structured extraction","family":"gpt-mini","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"benchmarks":[{"name":"Aider Polyglot","score":32.4,"metric":"percent correct","source":"https://aider.chat/docs/leaderboards/","date":"2025-04-14"}]},"openai/gpt-5.6-luna":{"id":"openai/gpt-5.6-luna","name":"GPT-5.6 Luna","description":"Cost-efficient GPT-5.6 model for fast, high-volume workloads","family":"gpt-luna","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"benchmarks":[{"name":"SWE-Bench Pro","score":62.7,"metric":"resolve rate","source":"https://openai.com/index/gpt-5-6/","date":"2026-07-09"},{"name":"Terminal-Bench","score":84.7,"metric":"success rate","version":"2.1","source":"https://openai.com/index/gpt-5-6/","date":"2026-07-09"},{"name":"DeepSWE","score":67.2,"metric":"resolve rate","version":"1.1","source":"https://openai.com/index/gpt-5-6/","date":"2026-07-09"},{"name":"GPQA Diamond","score":92.3,"metric":"accuracy","source":"https://openai.com/index/gpt-5-6/","date":"2026-07-09"},{"name":"FrontierMath","score":78.6,"metric":"accuracy","dataset":"Tier 1-3","version":"v2","source":"https://openai.com/index/gpt-5-6/","date":"2026-07-09"},{"name":"BrowseComp","score":83.3,"metric":"accuracy","source":"https://openai.com/index/gpt-5-6/","date":"2026-07-09"},{"name":"OSWorld","score":45.6,"metric":"success rate","version":"2.0","source":"https://openai.com/index/gpt-5-6/","date":"2026-07-09"},{"name":"MMMU Pro","score":78.4,"metric":"accuracy","variant":"no tools","source":"https://openai.com/index/gpt-5-6/","date":"2026-07-09"},{"name":"Agents' Last Exam","score":50.3,"source":"https://openai.com/index/gpt-5-6/","date":"2026-07-09"},{"name":"Toolathlon","score":53.4,"metric":"success rate","source":"https://openai.com/index/gpt-5-6/","date":"2026-07-09"},{"name":"Artificial Analysis Intelligence Index","score":51.2,"metric":"index score","variant":"max","version":"4.1","source":"https://artificialanalysis.ai/articles/gpt-5-6-has-landed","date":"2026-07-09"},{"name":"Artificial Analysis Coding Agent Index","score":74.6,"metric":"index score","harness":"Codex","variant":"max","version":"1.1","source":"https://artificialanalysis.ai/articles/gpt-5-6-has-landed","date":"2026-07-09"}]},"openai/gpt-5.6-sol":{"id":"openai/gpt-5.6-sol","name":"GPT-5.6 Sol","description":"Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows","family":"gpt-sol","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2026-02-16","release_date":"2026-07-09","last_updated":"2026-07-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"benchmarks":[{"name":"SWE-Bench Pro","score":64.6,"metric":"resolve rate","source":"https://openai.com/index/gpt-5-6/","date":"2026-07-09"},{"name":"Terminal-Bench","score":88.8,"metric":"success rate","version":"2.1","source":"https://openai.com/index/gpt-5-6/","date":"2026-07-09"},{"name":"DeepSWE","score":72.7,"metric":"resolve rate","version":"1.1","source":"https://openai.com/index/gpt-5-6/","date":"2026-07-09"},{"name":"GPQA Diamond","score":94.6,"metric":"accuracy","source":"https://openai.com/index/gpt-5-6/","date":"2026-07-09"},{"name":"FrontierMath","score":89,"metric":"accuracy","dataset":"Tier 1-3","version":"v2","source":"https://openai.com/index/gpt-5-6/","date":"2026-07-09"},{"name":"BrowseComp","score":90.4,"metric":"accuracy","source":"https://openai.com/index/gpt-5-6/","date":"2026-07-09"},{"name":"OSWorld","score":62.6,"metric":"success rate","version":"2.0","source":"https://openai.com/index/gpt-5-6/","date":"2026-07-09"},{"name":"MMMU Pro","score":83,"metric":"accuracy","variant":"no tools","source":"https://openai.com/index/gpt-5-6/","date":"2026-07-09"},{"name":"Agents' Last Exam","score":52.7,"source":"https://openai.com/index/gpt-5-6/","date":"2026-07-09"},{"name":"Toolathlon","score":58,"metric":"success rate","source":"https://openai.com/index/gpt-5-6/","date":"2026-07-09"},{"name":"Artificial Analysis Intelligence Index","score":58.9,"metric":"index score","variant":"max","version":"4.1","source":"https://artificialanalysis.ai/articles/gpt-5-6-has-landed","date":"2026-07-09"},{"name":"Artificial Analysis Coding Agent Index","score":80,"metric":"index score","harness":"Codex","variant":"max","version":"1.1","source":"https://artificialanalysis.ai/articles/gpt-5-6-has-landed","date":"2026-07-09"}]},"openai/gpt-image-1.5":{"id":"openai/gpt-image-1.5","type":"image","name":"GPT-Image-1.5","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-11-25","last_updated":"2025-11-25","modalities":{"input":["text","image"],"output":["text","image"]},"open_weights":false,"limit":{"context":0,"output":0}},"openai/gpt-5.2":{"id":"openai/gpt-5.2","name":"GPT-5.2","description":"Reliable GPT generation for broad coding, writing, and tool-assisted product work","family":"gpt","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"benchmarks":[{"name":"SWE-Bench Pro","score":29.94,"metric":"resolve rate","dataset":"public","source":"https://labs.scale.com/leaderboard/swe_bench_pro_public"}]},"openai/gpt-6-astra":{"id":"openai/gpt-6-astra","name":"GPT-6 Astra","description":"GPT-6 Astra is OpenAI's most capable model for complex reasoning, coding, computer use, research, and document creation.","family":"gpt-astra","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-04","last_updated":"2026-09-04","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"benchmarks":[{"name":"Agents' Last Exam","score":59.3,"source":"https://openai.com/index/gpt-6-astra/","date":"2026-09-03"},{"name":"OSWorld","score":72.6,"metric":"partial score","dataset":"V2-Offline (v2026.08.08)","version":"2.0","source":"https://openai.com/index/gpt-6-astra/","date":"2026-09-03"},{"name":"ScreenSpot-Pro","score":92.7,"metric":"accuracy","variant":"no tools","source":"https://openai.com/index/gpt-6-astra/","date":"2026-09-03"},{"name":"AutomationBench","score":41.4,"metric":"success rate","source":"https://openai.com/index/gpt-6-astra/","date":"2026-09-03"},{"name":"BenchCAD","score":95.9,"metric":"geometric overlap","variant":"with tools","source":"https://openai.com/index/gpt-6-astra/","date":"2026-09-03"},{"name":"BrowseComp","score":91.5,"metric":"accuracy","source":"https://openai.com/index/gpt-6-astra/","date":"2026-09-03"},{"name":"Terminal-Bench","score":57.9,"metric":"success rate","version":"4.0","source":"https://openai.com/index/gpt-6-astra/","date":"2026-09-03"},{"name":"DeepSWE","score":74.1,"metric":"resolve rate","version":"1.1","source":"https://openai.com/index/gpt-6-astra/","date":"2026-09-03"},{"name":"FrontierCode","score":64.5,"metric":"score","dataset":"Extended","version":"1.1","source":"https://openai.com/index/gpt-6-astra/","date":"2026-09-03"},{"name":"Terminal-Bench Science","score":64.6,"metric":"success rate","version":"0.1","source":"https://openai.com/index/gpt-6-astra/","date":"2026-09-03"},{"name":"FrontierMath","score":97.6,"metric":"accuracy","dataset":"Tier 4","version":"v2","source":"https://openai.com/index/gpt-6-astra/","date":"2026-09-03"},{"name":"GPQA Diamond","score":96,"metric":"accuracy","source":"https://openai.com/index/gpt-6-astra/","date":"2026-09-03"},{"name":"Humanity's Last Exam","score":57.2,"metric":"accuracy","variant":"with tools","source":"https://openai.com/index/gpt-6-astra/","date":"2026-09-03"},{"name":"ExploitBench","score":100,"metric":"success rate","variant":"without production safeguards","source":"https://openai.com/index/gpt-6-astra/","date":"2026-09-03"},{"name":"SRE-Bench","score":88,"metric":"success rate","variant":"single attempt","source":"https://openai.com/index/gpt-6-astra/","date":"2026-09-03"},{"name":"ARC-AGI-3","score":99.9,"metric":"RHAE","harness":"Responses API","source":"https://openai.com/index/gpt-6-astra/","date":"2026-09-03"},{"name":"Artificial Analysis Intelligence Index","score":61,"metric":"index score","variant":"max","version":"4.1.1","source":"https://artificialanalysis.ai/articles/benchmarking-gpt-6-astra","date":"2026-09-03"},{"name":"Artificial Analysis Coding Agent Index","score":67,"metric":"index score","harness":"Codex","variant":"max","version":"1.4","source":"https://artificialanalysis.ai/articles/benchmarking-gpt-6-astra","date":"2026-09-03"}]},"openai/gpt-4.1":{"id":"openai/gpt-4.1","name":"GPT-4.1","description":"Long-lived GPT workhorse for coding, instruction following, and production apps","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-04-14","last_updated":"2025-04-14","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1047576,"output":32768},"benchmarks":[{"name":"Aider Polyglot","score":52.4,"metric":"percent correct","source":"https://aider.chat/docs/leaderboards/","date":"2025-04-14"}]},"openai/o4-mini":{"id":"openai/o4-mini","name":"o4-mini","description":"Fast o-series model for compact reasoning, coding, and tool use","family":"o-mini","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-04-16","last_updated":"2025-04-16","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"benchmarks":[{"name":"Aider Polyglot","score":72,"metric":"percent correct","source":"https://aider.chat/docs/leaderboards/","date":"2025-04-16"}]},"openai/gpt-5.4":{"id":"openai/gpt-5.4","name":"GPT-5.4","description":"Agent-ready GPT for coding and computer-use workflows at a lower cost","family":"gpt","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-03-05","last_updated":"2026-03-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"benchmarks":[{"name":"SWE-Bench Pro","score":59.1,"metric":"resolve rate","dataset":"public","source":"https://labs.scale.com/leaderboard/swe_bench_pro_public"},{"name":"SWE-Atlas Codebase QnA","score":40.8,"metric":"score","harness":"Codex","source":"https://labs.scale.com/leaderboard/sweatlas-qna"},{"name":"SWE-Atlas Codebase QnA","score":36.3,"metric":"score","harness":"Mini-SWE-Agent","source":"https://labs.scale.com/leaderboard/sweatlas-qna"},{"name":"SWE-Atlas Refactoring","score":44.29,"metric":"score","harness":"Codex","source":"https://labs.scale.com/leaderboard/sweatlas-refactoring"},{"name":"SWE-Atlas Test Writing","score":44.36,"metric":"score","harness":"Codex CLI","source":"https://labs.scale.com/leaderboard/sweatlas-tw"},{"name":"SWE-Atlas Test Writing","score":40,"metric":"score","harness":"Mini-SWE-Agent","source":"https://labs.scale.com/leaderboard/sweatlas-tw"},{"name":"Artificial Analysis Coding Agent Index","score":53.6,"metric":"average pass@1","harness":"Codex","variant":"medium","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"SWE-Atlas Codebase QnA","score":72.4,"metric":"pass@1","harness":"Codex","variant":"medium","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"SWE-Bench Pro","score":18.4,"metric":"pass@1","harness":"Codex","variant":"medium","dataset":"hard-aa","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"Terminal-Bench","score":69.8,"metric":"pass@1","harness":"Codex","variant":"medium","version":"2.1","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"Artificial Analysis Coding Agent Index","score":52.2,"metric":"average pass@1","harness":"Cursor CLI","variant":"medium","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"SWE-Atlas Codebase QnA","score":72.9,"metric":"pass@1","harness":"Cursor CLI","variant":"medium","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"SWE-Bench Pro","score":18.9,"metric":"pass@1","harness":"Cursor CLI","variant":"medium","dataset":"hard-aa","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"Terminal-Bench","score":64.7,"metric":"pass@1","harness":"Cursor CLI","variant":"medium","version":"2.1","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"Terminal-Bench","score":75.1,"metric":"success rate","version":"2.0","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"GPQA Diamond","score":92.8,"metric":"accuracy","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"Humanity's Last Exam","score":39.8,"metric":"accuracy","variant":"no tools","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"Humanity's Last Exam","score":52.1,"metric":"accuracy","variant":"with tools","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"OSWorld-Verified","score":75,"metric":"success rate","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"BrowseComp","score":82.7,"metric":"accuracy","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"GDPval","score":83,"metric":"wins or ties","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"ARC-AGI-2","score":73.3,"metric":"accuracy","variant":"Verified","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"FrontierMath","score":47.6,"metric":"accuracy","dataset":"Tier 1-3","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"FrontierMath","score":27.1,"metric":"accuracy","dataset":"Tier 4","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"MMMU Pro","score":81.2,"metric":"accuracy","variant":"no tools","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"}]},"openai/gpt-oss-120b":{"id":"openai/gpt-oss-120b","name":"GPT OSS 120B","description":"Open GPT reasoning model for self-hosted agents and controllable deployments","family":"gpt-oss","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2024-06","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/openai/gpt-oss-120b"}]},"openai/gpt-5.1-codex-mini":{"id":"openai/gpt-5.1-codex-mini","name":"GPT-5.1 Codex mini","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-11-13","last_updated":"2025-11-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000}},"openai/gpt-image-1":{"id":"openai/gpt-image-1","type":"image","name":"GPT-Image-1","description":"OpenAI image model for production generation, edits, and brand-safe visual workflows","family":"gpt-image","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-04-24","last_updated":"2025-04-24","modalities":{"input":["text","image"],"output":["image"]},"open_weights":false,"limit":{"context":0,"output":0}},"openai/gpt-6.1-sol":{"id":"openai/gpt-6.1-sol","name":"GPT-6.1 Sol","description":"Near-Astra performance for complex coding, computer use, and professional work at a lower cost.","family":"gpt-sol","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2026-04-30","release_date":"2026-09-29","last_updated":"2026-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000}},"openai/gpt-5-image-mini":{"id":"openai/gpt-5-image-mini","type":"image","name":"GPT-5 Image Mini","description":"Image model for prompt-driven generation, editing, and visual design workflows","family":"gpt","attachment":true,"reasoning":true,"tool_call":false,"temperature":true,"release_date":"2025-10-16","last_updated":"2025-10-16","modalities":{"input":["pdf","image","text"],"output":["image","text"]},"open_weights":false,"limit":{"context":400000,"output":128000}},"openai/gpt-5-mini":{"id":"openai/gpt-5-mini","name":"GPT-5 Mini","description":"Small GPT-5 for responsive agents, coding help, and everyday automation","family":"gpt-mini","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2024-05-30","release_date":"2025-08-07","last_updated":"2025-08-07","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000}},"openai/o3-deep-research":{"id":"openai/o3-deep-research","name":"o3-deep-research","description":"Research model for long-horizon investigation, synthesis, and analytical reports","family":"o","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2024-05","release_date":"2025-06-26","last_updated":"2025-06-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000}},"openai/gpt-5.5-instant":{"id":"openai/gpt-5.5-instant","name":"GPT-5.5 Instant","description":"Compact GPT model for low-latency assistance and high-volume workloads","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-12-01","release_date":"2026-05-05","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000}},"openai/gpt-5-codex":{"id":"openai/gpt-5-codex","name":"GPT-5-Codex","description":"Coding-optimized GPT model for repository edits, reviews, and agentic software work","family":"gpt-codex","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2025-09-15","last_updated":"2025-09-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000},"benchmarks":[{"name":"Artificial Analysis Coding Index","score":38.9,"metric":"index","source":"https://openrouter.ai/openai/gpt-5-codex/benchmarks","date":"2026-06-01"},{"name":"SciCode","score":40.9,"metric":"percent correct","source":"https://openrouter.ai/openai/gpt-5-codex/benchmarks","date":"2026-06-01"},{"name":"Terminal-Bench Hard","score":37.9,"metric":"success rate","source":"https://openrouter.ai/openai/gpt-5-codex/benchmarks","date":"2026-06-01"}]},"openai/gpt-5.2-pro":{"id":"openai/gpt-5.2-pro","name":"GPT-5.2 Pro","description":"Higher-accuracy GPT-5.2 variant for tougher reasoning and review workflows","family":"gpt-pro","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2025-08-31","release_date":"2025-12-11","last_updated":"2025-12-11","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":400000,"input":272000,"output":128000}},"openai/gpt-realtime-2.1":{"id":"openai/gpt-realtime-2.1","type":"realtime","name":"GPT-Realtime-2.1","description":"Realtime speech-to-speech model with configurable reasoning, tool use, and robust voice-agent behavior","family":"gpt","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2024-09-30","release_date":"2026-07-06","last_updated":"2026-07-06","modalities":{"input":["text","audio","image"],"output":["text","audio"]},"open_weights":false,"limit":{"context":128000,"input":96000,"output":32000}},"openai/gpt-4o-2024-11-20":{"id":"openai/gpt-4o-2024-11-20","name":"GPT-4o (2024-11-20)","description":"GPT model for general reasoning, writing, coding, and tool-assisted tasks","family":"gpt","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-09","release_date":"2024-11-20","last_updated":"2024-11-20","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":16384},"benchmarks":[{"name":"Aider Polyglot","score":18.2,"metric":"percent correct","source":"https://aider.chat/docs/leaderboards/","date":"2024-12-30"},{"name":"Artificial Analysis Coding Index","score":16.7,"metric":"index","source":"https://openrouter.ai/openai/gpt-4o-2024-11-20/benchmarks","date":"2026-03-11"},{"name":"SciCode","score":33.3,"metric":"percent correct","source":"https://openrouter.ai/openai/gpt-4o-2024-11-20/benchmarks","date":"2026-03-11"},{"name":"Terminal-Bench Hard","score":8.3,"metric":"success rate","source":"https://openrouter.ai/openai/gpt-4o-2024-11-20/benchmarks","date":"2026-03-11"}]},"openai/o3-mini":{"id":"openai/o3-mini","name":"o3-mini","description":"Smaller o-series reasoner for economical coding, math, and planning tasks","family":"o-mini","attachment":false,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2024-05","release_date":"2024-12-20","last_updated":"2025-01-29","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":100000},"benchmarks":[{"name":"Aider Polyglot","score":60.4,"metric":"percent correct","source":"https://aider.chat/docs/leaderboards/","date":"2025-01-31"}]},"openai/gpt-6-luna":{"id":"openai/gpt-6-luna","name":"GPT-6 Luna","description":"OpenAI's most efficient model for focused, high-volume tasks","family":"gpt-luna","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2026-05-18","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1050000,"input":922000,"output":128000},"benchmarks":[{"name":"DeepSWE","score":66.6,"metric":"score","variant":"max effort","version":"1.1","source":"https://openai.com/index/introducing-gpt-6-sol-and-luna/","date":"2026-09-22"}]},"poolside/laguna-s-2.1":{"id":"poolside/laguna-s-2.1","name":"Laguna S 2.1","description":"Agentic coding model from Poolside in the XS size class for local deployment","family":"laguna","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-07-21","last_updated":"2026-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":32768}},"poolside/laguna-xs.2":{"id":"poolside/laguna-xs.2","name":"Laguna XS.2","description":"Agentic coding model from Poolside in the XS size class for local deployment","family":"laguna","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-04-28","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768}},"poolside/laguna-m.1":{"id":"poolside/laguna-m.1","name":"Laguna M.1","description":"Poolside's open-weight model for agentic coding and long-horizon work","family":"laguna","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-04-28","last_updated":"2026-06-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768}},"poolside/laguna-xs-2.1":{"id":"poolside/laguna-xs-2.1","name":"Laguna XS 2.1","description":"Agentic coding model from Poolside in the XS size class for local deployment","family":"laguna","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-07-02","last_updated":"2026-07-02","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"benchmarks":[{"name":"SWE-Bench Verified","score":70.9,"metric":"resolved","harness":"Harbor","source":"https://poolside.ai/blog/introducing-laguna-xs-2-1","date":"2026-07-02"},{"name":"SWE-Bench Multilingual","score":63.1,"metric":"resolve rate","harness":"Harbor","source":"https://poolside.ai/blog/introducing-laguna-xs-2-1","date":"2026-07-02"},{"name":"SWE-Bench Pro","score":47.6,"metric":"resolve rate","harness":"Harbor","source":"https://poolside.ai/blog/introducing-laguna-xs-2-1","date":"2026-07-02"},{"name":"Terminal-Bench","score":37.5,"metric":"success rate","harness":"Harbor","version":"2.0","source":"https://poolside.ai/blog/introducing-laguna-xs-2-1","date":"2026-07-02"}]},"recursive/flint":{"id":"recursive/flint","name":"Flint","description":"Open-weight mixture-of-experts model with about 2.5B active parameters for local coding agents","family":"flint","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2026-06","release_date":"2026-10-21","last_updated":"2026-10-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"license":"Apache-2.0","links":[{"label":"Model card","url":"https://huggingface.co/RecursiveSI/flint","type":"model_card"},{"label":"Introducing Recursive Flint","url":"https://www.recursive.com/updates/introducing-recursive-flint","type":"announcement"}],"weights":[{"label":"Hugging Face","url":"https://huggingface.co/RecursiveSI/flint"},{"label":"Hugging Face (INT4)","url":"https://huggingface.co/RecursiveSI/flint-int4","quantization":"int4"}]},"cohere/command-a-reasoning-08-2025":{"id":"cohere/command-a-reasoning-08-2025","name":"Command A Reasoning","description":"Cohere reasoning model for multilingual enterprise agents, tools, and complex workflows","family":"command-a","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":32000},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/CohereLabs/c4ai-command-a-reasoning-08-2025"}]},"cohere/command-a-plus-05-2026":{"id":"cohere/command-a-plus-05-2026","name":"Command A Plus","description":"Cohere's stronger command model for multilingual agents and enterprise workflows","family":"command-a","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-04-01","release_date":"2026-05-20","last_updated":"2026-06-09","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":64000}},"cohere/north-mini-code-1-0":{"id":"cohere/north-mini-code-1-0","name":"North Mini Code","description":"Cohere coding model for practical software engineering and agentic edits","family":"north","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-09-23","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":64000},"benchmarks":[{"name":"SWE-Bench Verified","score":67.6,"metric":"resolved","harness":"SWE-agent","source":"https://huggingface.co/CohereLabs/North-Mini-Code-1.0","date":"2026-06-09"},{"name":"SWE-Bench Pro","score":40.2,"metric":"resolve rate","harness":"SWE-agent","source":"https://huggingface.co/CohereLabs/North-Mini-Code-1.0","date":"2026-06-09"},{"name":"Artificial Analysis Intelligence Index","score":27.6,"metric":"index score","source":"https://artificialanalysis.ai/articles/north-mini-code-cohere-s-small-coding-focused-moe-model","date":"2026-06-09"},{"name":"Artificial Analysis Coding Index","score":33.4,"metric":"index score","source":"https://artificialanalysis.ai/articles/north-mini-code-cohere-s-small-coding-focused-moe-model","date":"2026-06-09"},{"name":"GDPval-AA","score":14,"metric":"win rate","source":"https://artificialanalysis.ai/articles/north-mini-code-cohere-s-small-coding-focused-moe-model","date":"2026-06-09"},{"name":"τ²-Bench Telecom","score":37,"metric":"success rate","source":"https://artificialanalysis.ai/articles/north-mini-code-cohere-s-small-coding-focused-moe-model","date":"2026-06-09"}]},"cohere/c4ai-aya-vision-8b":{"id":"cohere/c4ai-aya-vision-8b","name":"Aya Vision 8B","description":"Compact open multilingual vision model for OCR and visual question answering","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-03-04","last_updated":"2025-05-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":16000,"output":4000},"license":"CC-BY-NC-4.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/CohereLabs/aya-vision-8b"}]},"cohere/command-r7b-arabic-02-2025":{"id":"cohere/command-r7b-arabic-02-2025","name":"Command R7B Arabic","description":"Open Command R model optimized for Arabic enterprise chat, RAG, and cultural knowledge","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2025-02-27","last_updated":"2025-02-27","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/CohereLabs/c4ai-command-r7b-arabic-02-2025"}]},"cohere/tiny-aya-earth":{"id":"cohere/tiny-aya-earth","name":"Tiny Aya Earth","description":"Compact multilingual model specialized for West Asian and African languages","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-02-17","last_updated":"2026-02-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8000,"output":8000},"license":"CC-BY-NC-4.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/CohereLabs/tiny-aya-earth"}]},"cohere/command-r-plus-08-2024":{"id":"cohere/command-r-plus-08-2024","name":"Command R+","description":"Cohere's RAG workhorse for long-context enterprise search and tool use","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-08-30","last_updated":"2024-08-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/CohereLabs/c4ai-command-r-plus-08-2024"}]},"cohere/command-r7b-12-2024":{"id":"cohere/command-r7b-12-2024","name":"Command R7B","description":"Cohere retrieval model for long-context chat and enterprise RAG workflows","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-12-13","last_updated":"2024-12-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/CohereLabs/c4ai-command-r7b-12-2024"}]},"cohere/c4ai-aya-expanse-8b":{"id":"cohere/c4ai-aya-expanse-8b","name":"Aya Expanse 8B","description":"Compact open multilingual model optimized for generation across 23 languages","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2024-10-24","last_updated":"2024-10-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8000,"output":4000},"license":"CC-BY-NC-4.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/CohereLabs/aya-expanse-8b"}]},"cohere/north-small-translate-1-0":{"id":"cohere/north-small-translate-1-0","name":"North Small Translate","description":"Open-weight translation model for more than 50 languages and locale variants","family":"north","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-09-09","last_updated":"2026-09-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":16000,"output":16000},"license":"CC-BY-NC-4.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/CohereLabs/North-Small-Translate-1.0"}]},"cohere/command-a-translate-08-2025":{"id":"cohere/command-a-translate-08-2025","name":"Command A Translate","description":"Translation model for multilingual conversion, localization, and cross-language workflows","family":"command-a","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2025-08-28","last_updated":"2025-08-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8000,"output":8000},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/CohereLabs/c4ai-command-a-translate-08-2025"}]},"cohere/command-r-08-2024":{"id":"cohere/command-r-08-2024","name":"Command R","description":"Cohere retrieval model for long-context chat and enterprise RAG workflows","family":"command-r","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2024-08-30","last_updated":"2024-08-30","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/CohereLabs/c4ai-command-r-08-2024"}]},"cohere/tiny-aya-fire":{"id":"cohere/tiny-aya-fire","name":"Tiny Aya Fire","description":"Compact multilingual model specialized for South Asian languages","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-02-17","last_updated":"2026-02-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8000,"output":8000},"license":"CC-BY-NC-4.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/CohereLabs/tiny-aya-fire"}]},"cohere/c4ai-aya-expanse-32b":{"id":"cohere/c4ai-aya-expanse-32b","name":"Aya Expanse 32B","description":"Open multilingual model optimized for generation across 23 languages","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2024-10-24","last_updated":"2024-10-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":4000},"license":"CC-BY-NC-4.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/CohereLabs/aya-expanse-32b"}]},"cohere/command-a-vision-07-2025":{"id":"cohere/command-a-vision-07-2025","name":"Command A Vision","description":"Cohere vision model for multilingual document analysis, OCR, and image understanding","family":"command-a","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-06-01","release_date":"2025-07-31","last_updated":"2025-07-31","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8000},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/CohereLabs/c4ai-command-a-vision-07-2025"}]},"cohere/command-a-03-2025":{"id":"cohere/command-a-03-2025","name":"Command A","description":"Cohere command model for multilingual enterprise agents, tools, and chat","family":"command-a","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-06-01","release_date":"2025-03-13","last_updated":"2025-03-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":8000},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/CohereLabs/c4ai-command-a-03-2025"}],"benchmarks":[{"name":"Aider Polyglot","score":12,"metric":"percent correct","source":"https://aider.chat/docs/leaderboards/","date":"2025-03-14"}]},"cohere/command-a":{"id":"cohere/command-a","name":"Command A","description":"Cohere command model for multilingual enterprise agents, tools, and chat","family":"command-a","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-08-31","release_date":"2025-03-13","last_updated":"2025-03-13","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":256000,"output":8192}},"cohere/tiny-aya-water":{"id":"cohere/tiny-aya-water","name":"Tiny Aya Water","description":"Compact multilingual model specialized for European and Asia-Pacific languages","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-02-17","last_updated":"2026-02-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8000,"output":8000},"license":"CC-BY-NC-4.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/CohereLabs/tiny-aya-water"}]},"cohere/tiny-aya-global":{"id":"cohere/tiny-aya-global","name":"Tiny Aya Global","description":"Compact multilingual model balanced across 70 languages and regions","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-02-17","last_updated":"2026-02-17","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":8000,"output":8000},"license":"CC-BY-NC-4.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/CohereLabs/tiny-aya-global"}]},"cohere/c4ai-aya-vision-32b":{"id":"cohere/c4ai-aya-vision-32b","name":"Aya Vision 32B","description":"Open multilingual vision model for OCR, visual reasoning, and image question answering","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-03-04","last_updated":"2025-05-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":16000,"output":4000},"license":"CC-BY-NC-4.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/CohereLabs/aya-vision-32b"}]},"perplexity/sonar-deep-research":{"id":"perplexity/sonar-deep-research","name":"Sonar Deep Research","description":"Sonar search model for autonomous research and citation-backed long-form reports","family":"sonar","attachment":false,"reasoning":true,"tool_call":false,"temperature":false,"knowledge":"2025-01","release_date":"2025-02-01","last_updated":"2025-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":32768}},"perplexity/sonar":{"id":"perplexity/sonar","name":"Sonar","description":"Fast web-grounded Sonar for current answers, citations, and lightweight retrieval","family":"sonar","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-09-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096},"benchmarks":[{"name":"SciCode","score":22.9,"metric":"percent correct","source":"https://openrouter.ai/perplexity/sonar/benchmarks","date":"2026-03-11"}]},"perplexity/sonar-reasoning-pro":{"id":"perplexity/sonar-reasoning-pro","name":"Sonar Reasoning Pro","description":"Web-grounded Sonar for multi-step research questions that need cited reasoning","family":"sonar-reasoning","attachment":true,"reasoning":true,"tool_call":false,"temperature":true,"knowledge":"2025-09-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096}},"perplexity/sonar-pro":{"id":"perplexity/sonar-pro","name":"Sonar Pro","description":"Deeper Sonar search model with broader retrieval and stronger synthesis","family":"sonar-pro","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2025-09-01","release_date":"2024-01-01","last_updated":"2025-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8192},"benchmarks":[{"name":"SciCode","score":22.6,"metric":"percent correct","source":"https://openrouter.ai/perplexity/sonar-pro/benchmarks","date":"2026-03-11"}]},"alibaba/qwen-turbo":{"id":"alibaba/qwen-turbo","name":"Qwen Turbo","description":"Efficient Qwen model for fast chat, extraction, and high-volume workloads","family":"qwen","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-11-01","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":16384}},"alibaba/qwen3-next-80b-a3b-instruct":{"id":"alibaba/qwen3-next-80b-a3b-instruct","name":"Qwen3-Next 80B-A3B Instruct","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/Qwen/Qwen3-Next-80B-A3B-Instruct"}]},"alibaba/qwen3-235b-a22b-instruct-2507":{"id":"alibaba/qwen3-235b-a22b-instruct-2507","name":"Qwen3 235B-A22B Instruct 2507","description":"Updated large open Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-07-21","last_updated":"2025-07-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"license":"Apache 2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/Qwen/Qwen3-235B-A22B-Instruct-2507"}]},"alibaba/qwen3-omni-30b-a3b-thinking":{"id":"alibaba/qwen3-omni-30b-a3b-thinking","name":"Qwen3 Omni 30B A3B Thinking","description":"Multimodal reasoning model accepting text, images, video, and audio with text output","family":"qwen","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2025-09-22","last_updated":"2025-09-22","modalities":{"input":["text","image","video","audio"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":16384},"license":"Apache-2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/Qwen/Qwen3-Omni-30B-A3B-Thinking"}]},"alibaba/qwen3-embedding-8b":{"id":"alibaba/qwen3-embedding-8b","type":"embedding","name":"Qwen3 Embedding 8B","description":"Multilingual text embedding model for retrieval, code search, classification, and clustering","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":0},"license":"Apache-2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/Qwen/Qwen3-Embedding-8B"}]},"alibaba/qwen3.5-plus":{"id":"alibaba/qwen3.5-plus","name":"Qwen3.5 Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-02-16","last_updated":"2026-02-16","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536}},"alibaba/qwen-mt-plus":{"id":"alibaba/qwen-mt-plus","name":"Qwen MT Plus","description":"Flagship Qwen translation model with terminology control and support for 92 languages","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-07-24","last_updated":"2025-07-24","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":16384,"output":8192}},"alibaba/qwen3.5-35b-a3b":{"id":"alibaba/qwen3.5-35b-a3b","name":"Qwen3.5 35B-A3B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/Qwen/Qwen3.5-35B-A3B"}]},"alibaba/qwen3-235b-a22b":{"id":"alibaba/qwen3-235b-a22b","name":"Qwen3 235B-A22B","description":"Large open Qwen MoE for multilingual reasoning, coding, and tool use","family":"qwen","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/Qwen/Qwen3-235B-A22B"}],"benchmarks":[{"name":"Aider Polyglot","score":59.6,"metric":"percent correct","source":"https://aider.chat/docs/leaderboards/","date":"2025-05-09"},{"name":"SWE-Bench Pro","score":21.41,"metric":"resolve rate","dataset":"public","source":"https://labs.scale.com/leaderboard/swe_bench_pro_public"}]},"alibaba/qwen3-coder-30b-a3b-instruct":{"id":"alibaba/qwen3-coder-30b-a3b-instruct","name":"Qwen3-Coder 30B-A3B Instruct","description":"Smaller Qwen coder for efficient local agents and repo-level fixes","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-31","last_updated":"2025-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/Qwen/Qwen3-Coder-30B-A3B-Instruct"}],"benchmarks":[{"name":"Artificial Analysis Coding Index","score":19.4,"metric":"index","source":"https://openrouter.ai/qwen/qwen3-coder-30b-a3b-instruct/benchmarks","date":"2026-06-02"},{"name":"SciCode","score":27.8,"metric":"percent correct","source":"https://openrouter.ai/qwen/qwen3-coder-30b-a3b-instruct/benchmarks","date":"2026-06-02"},{"name":"Terminal-Bench Hard","score":15.2,"metric":"success rate","source":"https://openrouter.ai/qwen/qwen3-coder-30b-a3b-instruct/benchmarks","date":"2026-06-02"}]},"alibaba/qwen3.8-max":{"id":"alibaba/qwen3.8-max","name":"Qwen3.8 Max","description":"2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-08-03","last_updated":"2026-08-03","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"benchmarks":[{"name":"Terminal-Bench","score":86.6,"metric":"avg@10","harness":"Claude Code","variant":"5h timeout","version":"2.1","source":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B"},{"name":"SWE-Bench Pro","score":67.7,"metric":"resolved","harness":"Claude Code","dataset":"Qwen refined and corrected task set","source":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B"},{"name":"DeepSWE","score":56.6,"metric":"score","harness":"Claude Code","version":"1.1","source":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B"},{"name":"NL2Repo","score":55.9,"metric":"score","harness":"Claude Code","source":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B"},{"name":"FrontierSWE","score":73.5,"metric":"dominance score","source":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B","date":"2026-08-03"},{"name":"MLS-Bench-Lite","score":41,"metric":"score","harness":"Claude Code","variant":"5h timeout","source":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B"},{"name":"PaperBench","score":93,"metric":"score","harness":"BasicAgent","variant":"Code-Dev; 3 runs; 12h timeout; Opus 4.6 judge","source":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B"},{"name":"AndroidBench","score":75.1,"metric":"avg@3","dataset":"95 public tasks","source":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B"},{"name":"QwenSWEBench","score":80.7,"metric":"avg@3","harness":"Claude Code","dataset":"internal","source":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B"},{"name":"QwenQoderBench","score":58.4,"metric":"avg@5","harness":"Claude Code","dataset":"internal","source":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B"},{"name":"QwenReactBench","score":1724,"metric":"Elo","dataset":"internal","source":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B"},{"name":"QwenSVGBench","score":1713,"metric":"Elo","dataset":"internal","source":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B"},{"name":"CoWorkBench","score":74.8,"metric":"score","dataset":"internal","source":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B"},{"name":"WorkSpaceBench","score":67.7,"metric":"score","source":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B"},{"name":"JobBench","score":53.4,"metric":"score","source":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B"},{"name":"SkillsBench","score":70.2,"metric":"avg@3","harness":"OpenCode","dataset":"87 tasks","version":"1.1","source":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B"},{"name":"Agents' Last Exam","score":27,"metric":"pass@1","source":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B"},{"name":"Agents' Last Exam","score":52.4,"metric":"score","source":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B"},{"name":"AutomationBench","score":27.3,"metric":"pass@1","dataset":"600 public tasks","source":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B"},{"name":"Toolathlon-Verified","score":72.5,"metric":"pass@1","source":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B"},{"name":"WideSearch","score":81.9,"metric":"average item F1","harness":"Qwen-Agent","source":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B"},{"name":"Humanity's Last Exam","score":56.2,"metric":"score","variant":"with tools","source":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B"},{"name":"Humanity's Last Exam","score":43.6,"metric":"score","variant":"without tools","source":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B"},{"name":"GPQA Diamond","score":92.6,"metric":"score","source":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B"},{"name":"IFBench","score":82.8,"metric":"score","source":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B"},{"name":"$OneMillion-Bench","score":52.5,"metric":"expert score","variant":"gemini-3.1-pro-preview judge","source":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B"},{"name":"HealthBench","score":60.2,"metric":"score","source":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B"},{"name":"PLawBench","score":73.2,"metric":"score","variant":"gemini-3.1-pro-preview judge","source":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B"},{"name":"PRBench","score":57.6,"metric":"score","variant":"legal","source":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B"},{"name":"PRBench","score":58.3,"metric":"score","variant":"finance","source":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B"},{"name":"MRCR","score":92.9,"metric":"score","dataset":"256K; 8-needle","version":"2","source":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B"},{"name":"LongBench","score":66.3,"metric":"score","version":"2","source":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B"}]},"alibaba/qwen3.8-2.4t-a95b":{"id":"alibaba/qwen3.8-2.4t-a95b","name":"Qwen3.8 2.4T A95B","description":"Open-weight sparse MoE (2.4T total, 95B active), the open-weight twin of Qwen3.8 Max for coding, research, complex reasoning, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-08-12","last_updated":"2026-08-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"license":"qwen3.8-max","weights":[{"label":"Hugging Face","url":"https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B"}]},"alibaba/qwen-vl-max":{"id":"alibaba/qwen-vl-max","name":"Qwen-VL Max","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-04-08","last_updated":"2025-08-13","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":8192}},"alibaba/qwen-omni-turbo":{"id":"alibaba/qwen-omni-turbo","name":"Qwen-Omni Turbo","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-01-19","last_updated":"2025-03-26","modalities":{"input":["text","image","audio","video"],"output":["text","audio"]},"open_weights":false,"limit":{"context":32768,"output":2048}},"alibaba/qwq-32b":{"id":"alibaba/qwq-32b","name":"QwQ 32B","description":"Open reasoning model from the Qwen team for math, coding, and step-by-step problem solving","family":"qwen","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-03-05","last_updated":"2025-03-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/Qwen/QwQ-32B"}]},"alibaba/qwq-plus":{"id":"alibaba/qwq-plus","name":"QwQ Plus","description":"Qwen reasoning model for deliberate problem solving, math, and coding","family":"qwen","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-03-05","last_updated":"2025-03-05","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":8192}},"alibaba/qwen3.5-flash":{"id":"alibaba/qwen3.5-flash","name":"Qwen3.5 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536}},"alibaba/qwen2.5-coder-0.5b":{"id":"alibaba/qwen2.5-coder-0.5b","name":"Qwen2.5-Coder-0.5B","description":"Tiny open Qwen code model for lightweight completion and on-device coding","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2024-11-12","last_updated":"2024-11-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":8192},"license":"Apache 2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/Qwen/Qwen2.5-Coder-0.5B"}]},"alibaba/qwen-image-2.1":{"id":"alibaba/qwen-image-2.1","type":"image","name":"Qwen-Image-2.1","description":"Unified text-to-image generation and image editing model with native RGBA transparency and reference-image editing","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-14","last_updated":"2026-09-21","modalities":{"input":["text","image"],"output":["image"]},"open_weights":true,"limit":{"context":8192,"output":0},"license":"qwen-research","links":[{"label":"Blog","url":"https://qwen.ai/blog?id=qwen-image-2.1","type":"announcement"},{"label":"GitHub","url":"https://github.com/QwenLM/Qwen-Image-2.1","type":"docs"}],"weights":[{"label":"Hugging Face","url":"https://huggingface.co/Qwen/Qwen-Image-2.1"}]},"alibaba/qwen3.5-35b-a3b-base":{"id":"alibaba/qwen3.5-35b-a3b-base","name":"Qwen3.5 35B-A3B Base","description":"Pre-trained Qwen3.5 base checkpoint for fine-tuning and research, without instruction tuning","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-02-24","last_updated":"2026-02-24","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"license":"Apache-2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/Qwen/Qwen3.5-35B-A3B-Base"}]},"alibaba/qwen3-max-preview":{"id":"alibaba/qwen3-max-preview","name":"Qwen3 Max Preview","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-05","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768}},"alibaba/qwen3-30b-a3b-instruct-2507":{"id":"alibaba/qwen3-30b-a3b-instruct-2507","name":"Qwen3 30B-A3B Instruct 2507","description":"Updated compact Qwen3 MoE instruct model for multilingual chat, coding, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":16384},"license":"Apache 2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/Qwen/Qwen3-30B-A3B-Instruct-2507"}]},"alibaba/qwen3-next-80b-a3b-thinking":{"id":"alibaba/qwen3-next-80b-a3b-thinking","name":"Qwen3-Next 80B-A3B (Thinking)","description":"Efficient Qwen thinking model for local reasoning, math, and coding agents","family":"qwen","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09","last_updated":"2025-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/Qwen/Qwen3-Next-80B-A3B-Thinking"}]},"alibaba/qwen3.6-27b":{"id":"alibaba/qwen3.6-27b","name":"Qwen3.6 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/Qwen/Qwen3.6-27B"}],"benchmarks":[{"name":"SWE-Bench Verified","score":77.2,"metric":"resolved","source":"https://huggingface.co/Qwen/Qwen3.6-27B"}]},"alibaba/qwen3.5-9b-base":{"id":"alibaba/qwen3.5-9b-base","name":"Qwen3.5 9B Base","description":"Pre-trained Qwen3.5 base checkpoint for fine-tuning and research, without instruction tuning","family":"qwen","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-02-26","last_updated":"2026-02-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"license":"Apache-2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/Qwen/Qwen3.5-9B-Base"}]},"alibaba/qwen3-30b-a3b":{"id":"alibaba/qwen3-30b-a3b","name":"Qwen3 30B A3B","description":"Sparse MoE Qwen model with 3B active parameters for efficient chat and reasoning","family":"qwen","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/Qwen/Qwen3-30B-A3B"}]},"alibaba/qwen3.5-27b":{"id":"alibaba/qwen3.5-27b","name":"Qwen3.5 27B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/Qwen/Qwen3.5-27B"}],"benchmarks":[{"name":"SWE-Bench Verified","score":72.4,"metric":"resolved","source":"https://huggingface.co/Qwen/Qwen3.5-27B"}]},"alibaba/qwen3-vl-flash":{"id":"alibaba/qwen3-vl-flash","name":"Qwen3 VL Flash (Alibaba Cloud)","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768}},"alibaba/qwen3.5-2b":{"id":"alibaba/qwen3.5-2b","name":"Qwen3.5 2B","description":"Compact Qwen3.5 multimodal model for fast chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-03-02","last_updated":"2026-03-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":81920},"license":"Apache-2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/Qwen/Qwen3.5-2B"}]},"alibaba/qwen3.8-max-0902":{"id":"alibaba/qwen3.8-max-0902","name":"Qwen3.8 Max 0902","description":"2026-09-02 upgraded snapshot of Qwen3.8 Max with stronger coding, collaborative agents, and multimodal document understanding","family":"qwen","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-09-02","last_updated":"2026-09-02","modalities":{"input":["text","image","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072}},"alibaba/qwen3-8b":{"id":"alibaba/qwen3-8b","name":"Qwen3 8B","description":"Dense 8B Qwen3 model with switchable thinking for chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2025-04-29","last_updated":"2025-04-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":32768},"license":"Apache-2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/Qwen/Qwen3-8B"}]},"alibaba/qwen3-reranker-8b":{"id":"alibaba/qwen3-reranker-8b","type":"reranking","name":"Qwen3 Reranker 8B","description":"Multilingual query-document reranking model supporting more than 100 languages","family":"qwen","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2025-06-05","last_updated":"2025-06-05","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":32768,"output":0},"license":"Apache-2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/Qwen/Qwen3-Reranker-8B"}]},"alibaba/qwen3-coder-480b-a35b-instruct":{"id":"alibaba/qwen3-coder-480b-a35b-instruct","name":"Qwen3-Coder 480B-A35B Instruct","description":"Open Qwen coding heavyweight for repository reasoning and agentic engineering","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-22","last_updated":"2025-07-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/Qwen/Qwen3-Coder-480B-A35B-Instruct"}],"benchmarks":[{"name":"SWE-Bench Pro","score":38.7,"metric":"resolve rate","dataset":"public","source":"https://labs.scale.com/leaderboard/swe_bench_pro_public"}]},"alibaba/qwen2.5-7b-instruct":{"id":"alibaba/qwen2.5-7b-instruct","name":"Qwen2.5 7B Instruct","description":"Small open Qwen2.5 instruction model for chat, structured output, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2024-09-19","last_updated":"2024-09-19","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"license":"Apache 2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/Qwen/Qwen2.5-7B-Instruct"}]},"alibaba/qwen3-vl-235b-a22b-thinking":{"id":"alibaba/qwen3-vl-235b-a22b-thinking","name":"Qwen3 VL 235B A22B Thinking","description":"Qwen vision-language thinking model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/Qwen/Qwen3-VL-235B-A22B-Thinking"}]},"alibaba/qwen2-5-vl-72b-instruct":{"id":"alibaba/qwen2-5-vl-72b-instruct","name":"Qwen2.5-VL 72B Instruct","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-01-26","last_updated":"2025-01-26","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/Qwen/Qwen2.5-VL-72B-Instruct"}]},"alibaba/qwen3.5-4b":{"id":"alibaba/qwen3.5-4b","name":"Qwen3.5 4B","description":"Small Qwen3.5 multimodal model for low-cost chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-03-02","last_updated":"2026-03-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":81920},"license":"Apache-2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/Qwen/Qwen3.5-4B"}]},"alibaba/qwen3-max":{"id":"alibaba/qwen3-max","name":"Qwen3 Max","description":"Flagship Qwen3 model for coding agents, complex reasoning, and tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536},"benchmarks":[{"name":"Artificial Analysis Coding Index","score":26.4,"metric":"index","source":"https://openrouter.ai/qwen/qwen3-max/benchmarks","date":"2026-05-30"},{"name":"SciCode","score":38.3,"metric":"percent correct","source":"https://openrouter.ai/qwen/qwen3-max/benchmarks","date":"2026-05-30"},{"name":"Terminal-Bench Hard","score":20.5,"metric":"success rate","source":"https://openrouter.ai/qwen/qwen3-max/benchmarks","date":"2026-05-30"}]},"alibaba/qwen3-vl-30b-a3b-instruct":{"id":"alibaba/qwen3-vl-30b-a3b-instruct","name":"Qwen3 VL 30B A3B Instruct","description":"Open vision-language MoE model for image and video understanding, documents, and visual agents","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-10-04","last_updated":"2025-10-04","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"license":"Apache-2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/Qwen/Qwen3-VL-30B-A3B-Instruct"}]},"alibaba/qwen-vl-plus":{"id":"alibaba/qwen-vl-plus","name":"Qwen-VL Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-01-25","last_updated":"2025-08-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":131072,"output":8192}},"alibaba/qwen3.5-122b-a10b":{"id":"alibaba/qwen3.5-122b-a10b","name":"Qwen3.5 122B-A10B","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/Qwen/Qwen3.5-122B-A10B"}],"benchmarks":[{"name":"SWE-Bench Verified","score":72,"metric":"resolved","source":"https://huggingface.co/Qwen/Qwen3.5-122B-A10B"}]},"alibaba/qwen-max":{"id":"alibaba/qwen-max","name":"Qwen Max","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-04-03","last_updated":"2025-01-25","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":32768,"output":8192},"benchmarks":[{"name":"Aider Polyglot","score":21.8,"metric":"percent correct","source":"https://aider.chat/docs/leaderboards/","date":"2025-01-28"}]},"alibaba/qwen3-coder-plus":{"id":"alibaba/qwen3-coder-plus","name":"Qwen3 Coder Plus","description":"Hosted Qwen coder for software agents, repo edits, and long-context code","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-23","last_updated":"2025-07-23","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":65536}},"alibaba/qwen3-coder-next":{"id":"alibaba/qwen3-coder-next","name":"Qwen3 Coder Next","description":"Open-weight Qwen coding model for agents, repository edits, and multi-turn tool use","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2026-02-03","last_updated":"2026-02-03","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/Qwen/Qwen3-Coder-Next"}]},"alibaba/qwen3.8-max-preview":{"id":"alibaba/qwen3.8-max-preview","name":"Qwen3.8 Max Preview","description":"Preview Qwen flagship for million-token multimodal reasoning and long-horizon agentic workflows","family":"qwen","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-07-19","last_updated":"2026-07-19","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"benchmarks":[{"name":"Terminal-Bench","score":86.6,"metric":"accuracy","variant":"xhigh","version":"2.1","source":"https://www.alibabacloud.com/blog/qwen3-8-max-a-new-bar-for-coding-and-cowork_603421","date":"2026-08-03"},{"name":"SWE-Bench Pro","score":67.7,"metric":"resolve rate","harness":"Claude Code","variant":"xhigh","source":"https://www.alibabacloud.com/blog/qwen3-8-max-a-new-bar-for-coding-and-cowork_603421","date":"2026-08-03"},{"name":"DeepSWE","score":56.6,"metric":"resolve rate","harness":"Claude Code","variant":"xhigh","version":"1.1","source":"https://www.alibabacloud.com/blog/qwen3-8-max-a-new-bar-for-coding-and-cowork_603421","date":"2026-08-03"},{"name":"NL2Repo","score":55.9,"metric":"resolve rate","harness":"Claude Code","variant":"xhigh","source":"https://www.alibabacloud.com/blog/qwen3-8-max-a-new-bar-for-coding-and-cowork_603421","date":"2026-08-03"},{"name":"FrontierSWE","score":73.5,"metric":"dominance score","harness":"Claude Code","variant":"xhigh","source":"https://www.alibabacloud.com/blog/qwen3-8-max-a-new-bar-for-coding-and-cowork_603421","date":"2026-08-03"},{"name":"MLS-Bench-Lite","score":41,"metric":"score","harness":"Claude Code","variant":"xhigh","source":"https://www.alibabacloud.com/blog/qwen3-8-max-a-new-bar-for-coding-and-cowork_603421","date":"2026-08-03"},{"name":"AutomationBench","score":27.3,"metric":"pass@1","variant":"xhigh","dataset":"600-task public subset","source":"https://www.alibabacloud.com/blog/qwen3-8-max-a-new-bar-for-coding-and-cowork_603421","date":"2026-08-03"},{"name":"Toolathlon Verified","score":72.5,"metric":"pass@1","variant":"xhigh","source":"https://www.alibabacloud.com/blog/qwen3-8-max-a-new-bar-for-coding-and-cowork_603421","date":"2026-08-03"},{"name":"WideSearch","score":81.9,"metric":"F1","variant":"xhigh","source":"https://www.alibabacloud.com/blog/qwen3-8-max-a-new-bar-for-coding-and-cowork_603421","date":"2026-08-03"},{"name":"Humanity's Last Exam","score":56.2,"metric":"accuracy","variant":"xhigh, with tools","source":"https://www.alibabacloud.com/blog/qwen3-8-max-a-new-bar-for-coding-and-cowork_603421","date":"2026-08-03"},{"name":"GPQA Diamond","score":92.6,"metric":"accuracy","variant":"xhigh","source":"https://www.alibabacloud.com/blog/qwen3-8-max-a-new-bar-for-coding-and-cowork_603421","date":"2026-08-03"},{"name":"Humanity's Last Exam","score":43.6,"metric":"accuracy","variant":"xhigh, no tools","source":"https://www.alibabacloud.com/blog/qwen3-8-max-a-new-bar-for-coding-and-cowork_603421","date":"2026-08-03"},{"name":"IFBench","score":82.8,"metric":"score","variant":"xhigh","source":"https://www.alibabacloud.com/blog/qwen3-8-max-a-new-bar-for-coding-and-cowork_603421","date":"2026-08-03"},{"name":"OSWorld-Verified","score":86.1,"metric":"success rate","variant":"xhigh","source":"https://www.alibabacloud.com/blog/qwen3-8-max-a-new-bar-for-coding-and-cowork_603421","date":"2026-08-03"},{"name":"MMMU Pro","score":82.3,"metric":"accuracy","variant":"xhigh","source":"https://www.alibabacloud.com/blog/qwen3-8-max-a-new-bar-for-coding-and-cowork_603421","date":"2026-08-03"}]},"alibaba/qwen3.6-flash":{"id":"alibaba/qwen3.6-flash","name":"Qwen3.6 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen3.6","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-04-27","last_updated":"2026-04-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536}},"alibaba/qwen3.7-flash":{"id":"alibaba/qwen3.7-flash","name":"Qwen3.7 Flash","description":"Lightweight multimodal Qwen model for high-throughput text, image, and video tasks","family":"qwen","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-07-15","last_updated":"2026-07-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"input":991808,"output":131072}},"alibaba/qwen3.5-397b-a17b":{"id":"alibaba/qwen3.5-397b-a17b","name":"Qwen3.5 397B-A17B","description":"Large open Qwen multimodal MoE for visual agents and long technical tasks","family":"qwen","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-02-15","last_updated":"2026-02-15","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/Qwen/Qwen3.5-397B-A17B"}],"benchmarks":[{"name":"SWE-Bench Verified","score":76.4,"metric":"resolved","source":"https://huggingface.co/Qwen/Qwen3.5-397B-A17B"}]},"alibaba/qwen-flash":{"id":"alibaba/qwen-flash","name":"Qwen Flash","description":"Efficient Qwen model for fast chat, extraction, and high-volume workloads","family":"qwen","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768}},"alibaba/qwen3.7-max":{"id":"alibaba/qwen3.7-max","name":"Qwen3.7 Max","description":"Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks","family":"qwen","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-05-21","last_updated":"2026-05-21","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072},"benchmarks":[{"name":"SWE-Bench Verified","score":80.4,"metric":"resolved","source":"https://qwen.ai/blog?id=qwen3.7","date":"2026-05-19"},{"name":"SWE-Bench Pro","score":60.6,"metric":"resolve rate","source":"https://qwen.ai/blog?id=qwen3.7","date":"2026-05-19"},{"name":"SWE-Bench Multilingual","score":78.3,"metric":"resolve rate","source":"https://qwen.ai/blog?id=qwen3.7","date":"2026-05-19"},{"name":"Terminal-Bench","score":69.7,"metric":"success rate","harness":"Terminus-2","version":"2.0","source":"https://qwen.ai/blog?id=qwen3.7","date":"2026-05-19"},{"name":"GPQA Diamond","score":92.4,"metric":"accuracy","source":"https://qwen.ai/blog?id=qwen3.7","date":"2026-05-19"},{"name":"Humanity's Last Exam","score":41.4,"metric":"accuracy","source":"https://qwen.ai/blog?id=qwen3.7","date":"2026-05-19"},{"name":"SciCode","score":53.5,"source":"https://qwen.ai/blog?id=qwen3.7","date":"2026-05-19"},{"name":"MCP Atlas","score":76.4,"metric":"success rate","source":"https://qwen.ai/blog?id=qwen3.7","date":"2026-05-19"},{"name":"NL2Repo","score":47.2,"harness":"Claude Code","source":"https://qwen.ai/blog?id=qwen3.7","date":"2026-05-19"}]},"alibaba/qwen3.7-plus":{"id":"alibaba/qwen3.7-plus","name":"Qwen3.7 Plus","description":"Multimodal Qwen workhorse for long-context agents, visual inputs, and coding","family":"qwen","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-06-02","last_updated":"2026-06-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000}},"alibaba/qwen3-coder-flash":{"id":"alibaba/qwen3-coder-flash","name":"Qwen3 Coder Flash","description":"Qwen coding model for software agents, repository edits, and code reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-07-28","last_updated":"2025-07-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536}},"alibaba/qwen3-vl-235b-a22b-instruct":{"id":"alibaba/qwen3-vl-235b-a22b-instruct","name":"Qwen3 VL 235B A22B Instruct","description":"Qwen vision-language instruct model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/Qwen/Qwen3-VL-235B-A22B-Instruct"}]},"alibaba/qwen3.5-9b":{"id":"alibaba/qwen3.5-9b","name":"Qwen3.5 9B","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-02-23","last_updated":"2026-02-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/Qwen/Qwen3.5-9B"}]},"alibaba/qwen3.6-max-preview":{"id":"alibaba/qwen3.6-max-preview","name":"Qwen3.6 Max Preview","description":"Flagship Qwen model for complex reasoning, coding, and agentic workflows","family":"qwen","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-20","last_updated":"2026-04-20","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":65536}},"alibaba/qwen3-vl-plus":{"id":"alibaba/qwen3-vl-plus","name":"Qwen3-VL Plus","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-09-23","last_updated":"2025-09-23","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":32768}},"alibaba/qwen3.8-flash":{"id":"alibaba/qwen3.8-flash","name":"Qwen3.8 Flash","description":"Qwen vision-language model for visual reasoning, documents, and agent tasks","family":"qwen","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-08-26","last_updated":"2026-08-26","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072}},"alibaba/qwen3.8-omni-flash":{"id":"alibaba/qwen3.8-omni-flash","name":"Qwen3.8 Omni Flash","description":"Qwen omni model for text, vision, audio, and multimodal agent tasks","family":"qwen","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-09-17","last_updated":"2026-09-17","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072}},"alibaba/qwen3.8-27b":{"id":"alibaba/qwen3.8-27b","name":"Qwen3.8 27B","description":"Dense 27B vision-language model for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-08-14","last_updated":"2026-08-14","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":32768},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/Qwen/Qwen3.8-27B"}],"benchmarks":[{"name":"SWE-bench Pro","score":61.7,"metric":"resolved","source":"https://huggingface.co/Qwen/Qwen3.8-27B"},{"name":"Terminal-Bench","score":73,"metric":"score","harness":"Terminus","version":"2.1","source":"https://huggingface.co/Qwen/Qwen3.8-27B"},{"name":"NL2Repo","score":42.3,"metric":"score","harness":"Claude Code","source":"https://huggingface.co/Qwen/Qwen3.8-27B"},{"name":"DeepSWE","score":42.2,"metric":"score","harness":"Claude Code","version":"1.1","source":"https://huggingface.co/Qwen/Qwen3.8-27B"},{"name":"QwenSWEBench","score":79,"metric":"avg@3","harness":"Claude Code","dataset":"internal","source":"https://huggingface.co/Qwen/Qwen3.8-27B"},{"name":"CoWorkBench","score":70.7,"metric":"score","dataset":"internal","source":"https://huggingface.co/Qwen/Qwen3.8-27B"},{"name":"JobBench","score":33.4,"metric":"score","source":"https://huggingface.co/Qwen/Qwen3.8-27B"},{"name":"Agents' Last Exam","score":20.4,"metric":"pass@1","source":"https://huggingface.co/Qwen/Qwen3.8-27B"},{"name":"Agents' Last Exam","score":42.9,"metric":"score","source":"https://huggingface.co/Qwen/Qwen3.8-27B"},{"name":"IFBench","score":79.5,"metric":"score","source":"https://huggingface.co/Qwen/Qwen3.8-27B"},{"name":"GPQA Diamond","score":89.2,"metric":"score","source":"https://huggingface.co/Qwen/Qwen3.8-27B"},{"name":"Humanity's Last Exam","score":30.8,"metric":"score","variant":"without tools; GPT-4o judge","source":"https://huggingface.co/Qwen/Qwen3.8-27B"},{"name":"LiveCodeBench","score":90.3,"metric":"score","version":"6","source":"https://huggingface.co/Qwen/Qwen3.8-27B"},{"name":"OSWorld-Verified","score":84.3,"metric":"score","source":"https://huggingface.co/Qwen/Qwen3.8-27B"},{"name":"WebArena-Verified","score":64.8,"metric":"score","harness":"OSWorld scaffold; official WebArena-Verified grader","source":"https://huggingface.co/Qwen/Qwen3.8-27B"},{"name":"AndroidWorld","score":81.9,"metric":"score","source":"https://huggingface.co/Qwen/Qwen3.8-27B"},{"name":"SWE-MM","score":38.6,"metric":"score","source":"https://huggingface.co/Qwen/Qwen3.8-27B"},{"name":"RealWorldQA","score":85.9,"metric":"score","source":"https://huggingface.co/Qwen/Qwen3.8-27B"},{"name":"ERQA","score":65.5,"metric":"score","source":"https://huggingface.co/Qwen/Qwen3.8-27B"},{"name":"RecreationBench","score":47.1,"metric":"score","dataset":"internal","source":"https://huggingface.co/Qwen/Qwen3.8-27B"},{"name":"ClawEval-MM","score":57.4,"metric":"pass@3","source":"https://huggingface.co/Qwen/Qwen3.8-27B"},{"name":"ClawEval-MM","score":56.9,"metric":"average score","source":"https://huggingface.co/Qwen/Qwen3.8-27B"},{"name":"Vision2Web","score":62.9,"metric":"score","harness":"Claude Code","variant":"gpt-5.4-2026-03-05 judge","source":"https://huggingface.co/Qwen/Qwen3.8-27B"},{"name":"MathVision","score":90,"metric":"score","variant":"without code interpreter","dataset":"corrected annotations","source":"https://huggingface.co/Qwen/Qwen3.8-27B"},{"name":"MathVision","score":94.6,"metric":"score","variant":"with code interpreter","dataset":"corrected annotations","source":"https://huggingface.co/Qwen/Qwen3.8-27B"},{"name":"BabyVision","score":65.7,"metric":"score","variant":"without code interpreter","source":"https://huggingface.co/Qwen/Qwen3.8-27B"},{"name":"BabyVision","score":85.6,"metric":"score","variant":"with code interpreter","source":"https://huggingface.co/Qwen/Qwen3.8-27B"},{"name":"CharXiv Reasoning","score":83.7,"metric":"score","variant":"without code interpreter","dataset":"corrected annotations","source":"https://huggingface.co/Qwen/Qwen3.8-27B"},{"name":"CharXiv Reasoning","score":90.2,"metric":"score","variant":"with code interpreter","dataset":"corrected annotations","source":"https://huggingface.co/Qwen/Qwen3.8-27B"},{"name":"OmniDocBench","score":91.1,"metric":"score","version":"1.5","source":"https://huggingface.co/Qwen/Qwen3.8-27B"}]},"alibaba/qwen3.5-0.8b":{"id":"alibaba/qwen3.5-0.8b","name":"Qwen3.5 0.8B","description":"Smallest Qwen3.5 multimodal model for on-device chat, reasoning, and tool use","family":"qwen","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-03-02","last_updated":"2026-03-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":81920},"license":"Apache-2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/Qwen/Qwen3.5-0.8B"}]},"alibaba/qwen3.6-plus":{"id":"alibaba/qwen3.6-plus","name":"Qwen3.6 Plus","description":"Earlier Qwen multimodal workhorse for million-token agent and document tasks","family":"qwen","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2026-04-02","last_updated":"2026-04-02","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":65536}},"alibaba/qwen3.8-max-prime":{"id":"alibaba/qwen3.8-max-prime","name":"Qwen 3.8 Max Prime","description":"High-throughput edition of Qwen3.8 Max for coding, professional work, multimodal understanding, and long-running agent workflows","family":"qwen3.8-max","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-09-23","last_updated":"2026-09-23","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":131072}},"alibaba/qwen3-235b-a22b-thinking-2507":{"id":"alibaba/qwen3-235b-a22b-thinking-2507","name":"Qwen3 235B-A22B Thinking 2507","description":"Updated large open Qwen3 MoE thinking model for reasoning, math, coding, and tool use","family":"qwen","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2025-07-25","last_updated":"2025-07-25","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":81920},"license":"Apache 2.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/Qwen/Qwen3-235B-A22B-Thinking-2507"}]},"alibaba/qwen2.5-coder-32b-instruct":{"id":"alibaba/qwen2.5-coder-32b-instruct","name":"Qwen2.5-Coder-32B-Instruct","description":"Open coding-focused Qwen model for code generation, repair, and repository reasoning","family":"qwen","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2024-11-12","last_updated":"2024-11-12","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/Qwen/Qwen2.5-Coder-32B-Instruct"}]},"alibaba/qwen3.6-35b-a3b":{"id":"alibaba/qwen3.6-35b-a3b","name":"Qwen3.6 35B-A3B","description":"Open multimodal Qwen MoE for local agents that need vision, audio, and code","family":"qwen","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-04-17","last_updated":"2026-04-17","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/Qwen/Qwen3.6-35B-A3B"}],"benchmarks":[{"name":"SWE-Bench Verified","score":73.4,"metric":"resolved","source":"https://huggingface.co/Qwen/Qwen3.6-35B-A3B"}]},"alibaba/qwen3-32b":{"id":"alibaba/qwen3-32b","name":"Qwen3 32B","description":"Dense open Qwen model for self-hosted chat, reasoning, and coding","family":"qwen","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-04","release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":16384},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/Qwen/Qwen3-32B"}],"benchmarks":[{"name":"Aider Polyglot","score":40,"metric":"percent correct","source":"https://aider.chat/docs/leaderboards/","date":"2025-05-08"}]},"alibaba/qwen-plus":{"id":"alibaba/qwen-plus","name":"Qwen Plus","description":"Qwen instruction model for multilingual chat, reasoning, and tool use","family":"qwen","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2024-04","release_date":"2024-01-25","last_updated":"2025-09-11","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":32768}},"alibaba/qwen3.8-flash-next":{"id":"alibaba/qwen3.8-flash-next","name":"Qwen3.8 Flash Next","description":"Open-weight experimental preview of the Qwen4 architecture: hybrid-attention MoE (125B total, 6B active) with vision encoder for coding, agent tasks, and image and video understanding","family":"qwen","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-08-27","last_updated":"2026-08-27","modalities":{"input":["text","image","video"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":131072},"license":"qwen-community-1.0","weights":[{"label":"Hugging Face","url":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next"}],"benchmarks":[{"name":"DeepSWE","score":58.7,"metric":"score","harness":"mini-swe-agent","version":"1.1","source":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next"},{"name":"SWE-Bench Pro","score":62.5,"metric":"score","harness":"Claude Code","dataset":"Qwen refined and corrected task set","source":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next"},{"name":"SWE-Bench Multilingual","score":81,"metric":"score","harness":"mini-swe-agent","source":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next"},{"name":"NL2Repo","score":48.1,"metric":"score","harness":"Claude Code","source":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next"},{"name":"CoWorkBench","score":73.9,"metric":"score","dataset":"internal","source":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next"},{"name":"JobBench","score":55.7,"metric":"score","source":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next"},{"name":"Agents' Last Exam","score":24.3,"metric":"pass@1","source":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next"},{"name":"Agents' Last Exam","score":51.2,"metric":"score","source":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next"},{"name":"Toolathlon-Verified","score":73.5,"metric":"pass@1","source":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next"},{"name":"IFBench","score":81.3,"metric":"score","source":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next"},{"name":"GPQA Diamond","score":91.7,"metric":"score","source":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next"},{"name":"Humanity's Last Exam","score":35.9,"metric":"score","variant":"without tools; GPT-4o judge","source":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next"},{"name":"LiveCodeBench","score":91.9,"metric":"score","version":"6","source":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next"},{"name":"ClawEval-MM","score":64.4,"metric":"pass@3","source":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next"},{"name":"ClawEval-MM","score":60.4,"metric":"average score","source":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next"},{"name":"RecreationBench","score":49.9,"metric":"score","dataset":"internal","source":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next"},{"name":"AndroidWorld","score":84.5,"metric":"score","source":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next"},{"name":"OSWorld","score":19.4,"metric":"binary completion rate","version":"2.0","source":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next"},{"name":"OSWorld","score":52.3,"metric":"partial score","version":"2.0","source":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next"},{"name":"Vision2Web","score":64,"metric":"score","harness":"Claude Code","variant":"gpt-5.4-2026-03-05 judge","source":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next"},{"name":"ERQA","score":72.3,"metric":"score","source":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next"},{"name":"LVBench","score":76.6,"metric":"score","source":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next"},{"name":"RealWorldQA","score":88.5,"metric":"score","source":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next"},{"name":"MathVision","score":90.6,"metric":"score","variant":"without code interpreter","dataset":"corrected annotations","source":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next"},{"name":"MathVision","score":95.7,"metric":"score","variant":"with code interpreter","dataset":"corrected annotations","source":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next"},{"name":"CharXiv Reasoning","score":84.6,"metric":"score","variant":"without code interpreter","source":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next"},{"name":"CharXiv Reasoning","score":90.6,"metric":"score","variant":"with code interpreter","source":"https://huggingface.co/Qwen/Qwen3.8-Flash-Next"}]},"deepseek/deepseek-v4-flash":{"id":"deepseek/deepseek-v4-flash","name":"DeepSeek V4 Flash","description":"Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work","family":"deepseek-flash","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash"}],"benchmarks":[{"name":"SWE-Bench Verified","score":79,"metric":"resolved","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash"},{"name":"MMLU-Pro","score":86.2,"metric":"EM","variant":"preview checkpoint; max effort","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash"},{"name":"SimpleQA-Verified","score":34.1,"metric":"pass@1","variant":"preview checkpoint; max effort","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash"},{"name":"Chinese SimpleQA","score":78.9,"metric":"pass@1","variant":"preview checkpoint; max effort","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash"},{"name":"GPQA Diamond","score":88.1,"metric":"pass@1","variant":"preview checkpoint; max effort","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash"},{"name":"Humanity's Last Exam","score":34.8,"metric":"pass@1","variant":"preview checkpoint; max effort; without tools","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash"},{"name":"LiveCodeBench","score":91.6,"metric":"pass@1","variant":"preview checkpoint; max effort","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash"},{"name":"Codeforces","score":3052,"metric":"rating","variant":"preview checkpoint; max effort","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash"},{"name":"HMMT","score":94.8,"metric":"pass@1","variant":"preview checkpoint; max effort","dataset":"February 2026","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash"},{"name":"IMOAnswerBench","score":88.4,"metric":"pass@1","variant":"preview checkpoint; max effort","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash"},{"name":"MathArena Apex","score":33,"metric":"pass@1","variant":"preview checkpoint; max effort","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash"},{"name":"MathArena Apex Shortlist","score":85.7,"metric":"pass@1","variant":"preview checkpoint; max effort","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash"},{"name":"MRCR","score":78.7,"metric":"MMR","variant":"preview checkpoint; max effort","dataset":"1M context","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash"},{"name":"CorpusQA","score":60.5,"metric":"accuracy","variant":"preview checkpoint; max effort","dataset":"1M context","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash"},{"name":"Terminal-Bench","score":56.9,"metric":"accuracy","variant":"preview checkpoint; max effort","version":"2.0","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash"},{"name":"SWE-Bench Pro","score":52.6,"metric":"resolved","variant":"preview checkpoint; max effort","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash"},{"name":"SWE-Bench Multilingual","score":73.3,"metric":"resolved","variant":"preview checkpoint; max effort","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash"},{"name":"BrowseComp","score":73.2,"metric":"pass@1","variant":"preview checkpoint; max effort","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash"},{"name":"Humanity's Last Exam","score":45.1,"metric":"pass@1","variant":"preview checkpoint; max effort; with tools","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash"},{"name":"MCP Atlas","score":69,"metric":"pass@1","variant":"preview checkpoint; max effort","dataset":"public","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash"},{"name":"GDPval-AA","score":1395,"metric":"Elo","variant":"preview checkpoint; max effort","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash"},{"name":"Toolathlon","score":47.8,"metric":"pass@1","variant":"preview checkpoint; max effort","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash"}]},"deepseek/deepseek-v3":{"id":"deepseek/deepseek-v3","name":"DeepSeek-V3","description":"Open DeepSeek MoE chat model for coding, math, and general reasoning","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2024-12-26","last_updated":"2024-12-26","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"license":"DeepSeek Model License","weights":[{"label":"Hugging Face","url":"https://huggingface.co/deepseek-ai/DeepSeek-V3"}]},"deepseek/deepseek-r1-distill-qwen-32b":{"id":"deepseek/deepseek-r1-distill-qwen-32b","name":"DeepSeek-R1-Distill-Qwen-32B","description":"R1 reasoning distilled into Qwen 2.5 32B for efficient open-weight step-by-step problem solving","family":"deepseek-thinking","attachment":false,"reasoning":true,"tool_call":false,"temperature":true,"release_date":"2025-01-20","last_updated":"2025-01-20","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32768},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-32B"}]},"deepseek/deepseek-reasoner":{"id":"deepseek/deepseek-reasoner","name":"DeepSeek Reasoner","description":"DeepSeek reasoning model for multi-step analysis, math, coding, and tools","family":"deepseek-thinking","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-12-01","last_updated":"2026-02-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/deepseek-ai/DeepSeek-V3.2"}],"benchmarks":[{"name":"Aider Polyglot","score":74.2,"metric":"percent correct","source":"https://aider.chat/docs/leaderboards/","date":"2025-10-03"}]},"deepseek/deepseek-v4-pro-0813":{"id":"deepseek/deepseek-v4-pro-0813","name":"DeepSeek V4 Pro 0813","description":"DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-08-13","last_updated":"2026-08-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"license":"MIT","weights":[{"label":"Hugging Face","url":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813"}],"benchmarks":[{"name":"Humanity's Last Exam","score":42.7,"metric":"score","variant":"without tools","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813"},{"name":"Humanity's Last Exam","score":60,"metric":"score","variant":"with tools","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813"},{"name":"Terminal-Bench","score":87.9,"metric":"score","harness":"DeepSeek Harness minimal mode","variant":"max effort","version":"2.1","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813"},{"name":"NL2Repo","score":61.5,"metric":"score","harness":"DeepSeek Harness minimal mode","variant":"max effort","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813"},{"name":"CyberGym","score":83.3,"metric":"score","harness":"DeepSeek Harness minimal mode","variant":"max effort","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813"},{"name":"DeepSWE","score":62.7,"metric":"score","harness":"DeepSeek Harness minimal mode","variant":"max effort","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813"},{"name":"Toolathlon-Verified","score":74.1,"metric":"score","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813"},{"name":"Agents' Last Exam","score":25.7,"metric":"score","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813"},{"name":"AutomationBench","score":31.8,"metric":"score","dataset":"public","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813"},{"name":"DSBench-FullStack","score":71.1,"metric":"score","dataset":"internal","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813"},{"name":"DSBench-Hard","score":67.2,"metric":"score","dataset":"internal","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813"}]},"deepseek/deepseek-v3-0324":{"id":"deepseek/deepseek-v3-0324","name":"DeepSeek V3 0324","description":"March 2025 checkpoint of DeepSeek-V3 with improved reasoning and coding","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2025-03-24","last_updated":"2025-03-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":163840},"weights":[{"label":"Model weights","url":"https://huggingface.co/deepseek-ai/DeepSeek-V3-0324","format":"safetensors"}]},"deepseek/deepseek-r1-0528-qwen3-8b":{"id":"deepseek/deepseek-r1-0528-qwen3-8b","name":"DeepSeek R1 0528 Qwen3 8B","description":"DeepSeek R1 0528 distilled into Qwen3 8B for mathematics, coding, and reasoning","family":"deepseek-thinking","attachment":false,"reasoning":true,"tool_call":false,"temperature":true,"release_date":"2025-05-29","last_updated":"2025-05-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":32000},"license":"MIT","weights":[{"label":"Hugging Face","url":"https://huggingface.co/deepseek-ai/DeepSeek-R1-0528-Qwen3-8B"}]},"deepseek/deepseek-v4-flash-0731":{"id":"deepseek/deepseek-v4-flash-0731","name":"DeepSeek V4 Flash 0731","description":"Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding","family":"deepseek-flash","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-07-31","last_updated":"2026-07-31","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"license":"MIT","weights":[{"label":"Hugging Face","url":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731"}],"benchmarks":[{"name":"Terminal-Bench","score":82.7,"metric":"pass@1","variant":"max","version":"2.1","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731"},{"name":"NL2Repo","score":54.2,"metric":"resolve rate","harness":"DeepSeek Harness minimal mode","variant":"max effort","source":"https://api-docs.deepseek.com/updates/","date":"2026-07-31"},{"name":"CyberGym","score":76.7,"metric":"score","harness":"DeepSeek Harness minimal mode","variant":"max effort","source":"https://api-docs.deepseek.com/updates/","date":"2026-07-31"},{"name":"DeepSWE","score":54.4,"metric":"resolve rate","harness":"DeepSeek Harness minimal mode","variant":"max effort","source":"https://api-docs.deepseek.com/updates/","date":"2026-07-31"},{"name":"Toolathlon-Verified","score":70.3,"metric":"score","harness":"DeepSeek Harness minimal mode","variant":"max effort","source":"https://api-docs.deepseek.com/updates/","date":"2026-07-31"},{"name":"Agents' Last Exam","score":25.2,"metric":"score","harness":"DeepSeek Harness minimal mode","variant":"max effort","source":"https://api-docs.deepseek.com/updates/","date":"2026-07-31"},{"name":"AutomationBench","score":25.1,"metric":"success rate","variant":"max effort","dataset":"public","source":"https://api-docs.deepseek.com/updates/","date":"2026-07-31"},{"name":"DSBench-FullStack","score":68.7,"metric":"score","variant":"max effort","dataset":"internal","source":"https://api-docs.deepseek.com/updates/","date":"2026-07-31"},{"name":"DSBench-Hard","score":59.6,"metric":"score","variant":"max effort","dataset":"internal","source":"https://api-docs.deepseek.com/updates/","date":"2026-07-31"}]},"deepseek/deepseek-chat-v3-0324":{"id":"deepseek/deepseek-chat-v3-0324","name":"DeepSeek V3 0324","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-07-31","release_date":"2025-03-24","last_updated":"2025-03-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":147456}},"deepseek/deepseek-v4-flash-0423":{"id":"deepseek/deepseek-v4-flash-0423","name":"DeepSeek V4 Flash 0423","description":"Initial DeepSeek V4 Flash snapshot for economical reasoning, coding, and million-token agent workloads","family":"deepseek-flash","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"license":"MIT","weights":[{"label":"Hugging Face","url":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash"}]},"deepseek/deepseek-r1":{"id":"deepseek/deepseek-r1","name":"DeepSeek-R1","description":"Classic open reasoning model for transparent math, coding, and deliberate problem solving","family":"deepseek-thinking","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-01-20","last_updated":"2025-05-29","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":32768},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/deepseek-ai/DeepSeek-R1"}],"benchmarks":[{"name":"Aider Polyglot","score":56.9,"metric":"percent correct","source":"https://aider.chat/docs/leaderboards/","date":"2025-01-20"},{"name":"Artificial Analysis Coding Index","score":15.9,"metric":"index","source":"https://openrouter.ai/deepseek/deepseek-r1/benchmarks","date":"2026-03-11"},{"name":"SciCode","score":35.7,"metric":"percent correct","source":"https://openrouter.ai/deepseek/deepseek-r1/benchmarks","date":"2026-03-11"},{"name":"Terminal-Bench Hard","score":6.1,"metric":"success rate","source":"https://openrouter.ai/deepseek/deepseek-r1/benchmarks","date":"2026-03-11"}]},"deepseek/deepseek-v3.1":{"id":"deepseek/deepseek-v3.1","name":"DeepSeek-V3.1","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes","family":"deepseek","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":131072,"output":8192},"license":"MIT License","weights":[{"label":"Hugging Face","url":"https://huggingface.co/deepseek-ai/DeepSeek-V3.1"}]},"deepseek/deepseek-v4-pro":{"id":"deepseek/deepseek-v4-pro","name":"DeepSeek V4 Pro","description":"Open MoE flagship with million-token context for coding and long agent runs","family":"deepseek-thinking","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-24","last_updated":"2026-04-24","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro"}],"benchmarks":[{"name":"SWE-Bench Verified","score":80.6,"metric":"resolved","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro"},{"name":"Artificial Analysis Coding Agent Index","score":50.1,"metric":"average pass@1","harness":"Claude Code","variant":"high","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"SWE-Atlas Codebase QnA","score":67.8,"metric":"pass@1","harness":"Claude Code","variant":"high","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"SWE-Bench Pro","score":18,"metric":"pass@1","harness":"Claude Code","variant":"high","dataset":"hard-aa","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"Terminal-Bench","score":64.7,"metric":"pass@1","harness":"Claude Code","variant":"high","version":"2.1","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"MMLU-Pro","score":87.5,"metric":"EM","variant":"preview checkpoint; max effort","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro"},{"name":"SimpleQA-Verified","score":57.9,"metric":"pass@1","variant":"preview checkpoint; max effort","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro"},{"name":"Chinese SimpleQA","score":84.4,"metric":"pass@1","variant":"preview checkpoint; max effort","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro"},{"name":"GPQA Diamond","score":90.1,"metric":"pass@1","variant":"preview checkpoint; max effort","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro"},{"name":"Humanity's Last Exam","score":37.7,"metric":"pass@1","variant":"preview checkpoint; max effort; without tools","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro"},{"name":"LiveCodeBench","score":93.5,"metric":"pass@1","variant":"preview checkpoint; max effort","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro"},{"name":"Codeforces","score":3206,"metric":"rating","variant":"preview checkpoint; max effort","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro"},{"name":"HMMT","score":95.2,"metric":"pass@1","variant":"preview checkpoint; max effort","dataset":"February 2026","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro"},{"name":"IMOAnswerBench","score":89.8,"metric":"pass@1","variant":"preview checkpoint; max effort","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro"},{"name":"MathArena Apex","score":38.3,"metric":"pass@1","variant":"preview checkpoint; max effort","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro"},{"name":"MathArena Apex Shortlist","score":90.2,"metric":"pass@1","variant":"preview checkpoint; max effort","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro"},{"name":"MRCR","score":83.5,"metric":"MMR","variant":"preview checkpoint; max effort","dataset":"1M context","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro"},{"name":"CorpusQA","score":62,"metric":"accuracy","variant":"preview checkpoint; max effort","dataset":"1M context","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro"},{"name":"Terminal-Bench","score":67.9,"metric":"accuracy","variant":"preview checkpoint; max effort","version":"2.0","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro"},{"name":"SWE-Bench Pro","score":55.4,"metric":"resolved","variant":"preview checkpoint; max effort","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro"},{"name":"SWE-Bench Multilingual","score":76.2,"metric":"resolved","variant":"preview checkpoint; max effort","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro"},{"name":"BrowseComp","score":83.4,"metric":"pass@1","variant":"preview checkpoint; max effort","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro"},{"name":"Humanity's Last Exam","score":48.2,"metric":"pass@1","variant":"preview checkpoint; max effort; with tools","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro"},{"name":"MCP Atlas","score":73.6,"metric":"pass@1","variant":"preview checkpoint; max effort","dataset":"public","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro"},{"name":"GDPval-AA","score":1554,"metric":"Elo","variant":"preview checkpoint; max effort","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro"},{"name":"Toolathlon","score":51.8,"metric":"pass@1","variant":"preview checkpoint; max effort","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro"}]},"deepseek/deepseek-v4-flash-vision-exp":{"id":"deepseek/deepseek-v4-flash-vision-exp","name":"DeepSeek V4 Flash Vision Exp","description":"Experimental multimodal DeepSeek V4 Flash model for image understanding, coding, and agentic work","family":"deepseek-flash","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-08-21","last_updated":"2026-09-01","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"license":"MIT","weights":[{"label":"Hugging Face","url":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-Vision-Exp"}],"benchmarks":[{"name":"Terminal-Bench","score":83.9,"metric":"score","harness":"DeepSeek Harness minimal mode","variant":"max effort","version":"2.1","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-Vision-Exp"},{"name":"NL2Repo","score":57.7,"metric":"score","harness":"DeepSeek Harness minimal mode","variant":"max effort","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-Vision-Exp"},{"name":"CyberGym","score":75.3,"metric":"score","harness":"DeepSeek Harness minimal mode","variant":"max effort","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-Vision-Exp"},{"name":"DeepSWE","score":59.3,"metric":"score","harness":"DeepSeek Harness minimal mode","variant":"max effort","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-Vision-Exp"},{"name":"Toolathlon-Verified","score":75.9,"metric":"score","harness":"DeepSeek Harness minimal mode","variant":"max effort","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-Vision-Exp"},{"name":"DSBench-Hard","score":63.6,"metric":"score","harness":"DeepSeek Harness minimal mode","variant":"max effort","dataset":"internal","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-Vision-Exp"},{"name":"AutomationBench","score":25.7,"metric":"score","harness":"DeepSeek Harness minimal mode","variant":"max effort","dataset":"public","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-Vision-Exp"},{"name":"ApexBench","score":36.5,"metric":"pass@1","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-Vision-Exp"},{"name":"Agents' Last Exam","score":27.3,"metric":"score","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-Vision-Exp"},{"name":"Chartography","score":64.3,"metric":"score","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-Vision-Exp"},{"name":"ZeroBench","score":35,"metric":"pass@5","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-Vision-Exp"}]},"deepseek/deepseek-v4.1-flash":{"id":"deepseek/deepseek-v4.1-flash","name":"DeepSeek V4.1 Flash","description":"DeepSeek V4.1 Flash model for reasoning and agentic coding","family":"deepseek-flash","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-09-10","last_updated":"2026-09-10","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"license":"MIT","weights":[{"label":"Hugging Face","url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash"}],"benchmarks":[{"name":"GPQA Diamond","score":90.9,"metric":"pass@1","variant":"reasoning_effort=100","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash"},{"name":"MathArena Apex","score":65.6,"metric":"pass@1","variant":"reasoning_effort=100","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash"},{"name":"Humanity's Last Exam","score":36.8,"metric":"pass@1","variant":"reasoning_effort=100","dataset":"full set","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash"},{"name":"Humanity's Last Exam","score":39.1,"metric":"pass@1","variant":"reasoning_effort=100","dataset":"text-only subset","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash"},{"name":"Codeforces","score":3471,"metric":"rating","variant":"reasoning_effort=100","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash"},{"name":"Terminal-Bench","score":90.6,"metric":"pass@1","harness":"DeepSeek Harness minimal mode","variant":"reasoning_effort=100","version":"2.1","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash"},{"name":"Terminal-Bench","score":30,"metric":"pass@1","harness":"DeepSeek Harness minimal mode","variant":"reasoning_effort=100","version":"3.0","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash"},{"name":"Terminal-Bench","score":31.2,"metric":"pass@1","harness":"DeepSeek Harness minimal mode","variant":"reasoning_effort=100","version":"4.0","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash"},{"name":"DeepSWE","score":74.2,"metric":"resolved","harness":"mini-swe-agent","variant":"reasoning_effort=100","version":"1.1","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash"},{"name":"ProgramBench","score":20.3,"metric":"almost@1","harness":"DeepSeek Harness minimal mode","variant":"reasoning_effort=100","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash"},{"name":"NL2Repo","score":64,"metric":"score","harness":"DeepSeek Harness minimal mode","variant":"reasoning_effort=100","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash"},{"name":"CyberGym","score":88.1,"metric":"pass@1","variant":"reasoning_effort=100","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash"},{"name":"ExploitGym","score":15.3,"metric":"pass@1","variant":"reasoning_effort=100","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash"},{"name":"AutomationBench","score":54.8,"metric":"pass@1","variant":"reasoning_effort=100","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash"},{"name":"Agents' Last Exam","score":31.8,"metric":"pass@1","variant":"reasoning_effort=100","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash"},{"name":"SEC-Bench Pro","score":62.8,"metric":"pass@1","harness":"Claude Code","variant":"reasoning_effort=100","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash"},{"name":"Humanity's Last Exam","score":63.9,"metric":"pass@1","variant":"reasoning_effort=100; with tools","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash"},{"name":"Chartography","score":78.9,"metric":"pass@1","harness":"Claude Code","variant":"reasoning_effort=100; with tools","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash"},{"name":"BabyVision","score":89.6,"metric":"pass@1","harness":"Claude Code","variant":"reasoning_effort=100; with tools","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash"},{"name":"ZeroBench","score":49,"metric":"pass@5","harness":"Claude Code","variant":"reasoning_effort=100; with tools","dataset":"main","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash"},{"name":"DeepSWE","score":69.8,"metric":"resolved","harness":"Claude Code","variant":"reasoning_effort=100","version":"1.1","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash"},{"name":"Terminal-Bench","score":88,"metric":"pass@1","harness":"Claude Code","variant":"reasoning_effort=100","version":"2.1","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash"},{"name":"DeepSWE","score":65.6,"metric":"resolved","harness":"Codex","variant":"reasoning_effort=100","version":"1.1","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash"},{"name":"Terminal-Bench","score":84.1,"metric":"pass@1","harness":"Codex","variant":"reasoning_effort=100","version":"2.1","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash"},{"name":"DeepSWE","score":65.5,"metric":"resolved","harness":"OpenCode","variant":"reasoning_effort=100","version":"1.1","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash"},{"name":"Terminal-Bench","score":85,"metric":"pass@1","harness":"OpenCode","variant":"reasoning_effort=100","version":"2.1","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash"},{"name":"DeepSWE","score":66.2,"metric":"resolved","harness":"Pi","variant":"reasoning_effort=100","version":"1.1","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash"},{"name":"Terminal-Bench","score":86.1,"metric":"pass@1","harness":"Pi","variant":"reasoning_effort=100","version":"2.1","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash"},{"name":"Terminal-Bench","score":90.3,"metric":"pass@1","harness":"mini-swe-agent","variant":"reasoning_effort=100","version":"2.1","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash"},{"name":"DeepSWE","score":72.6,"metric":"resolved","harness":"DeepSeek Harness minimal mode","variant":"reasoning_effort=100","version":"1.1","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash"},{"name":"DeepSWE","score":70.5,"metric":"resolved","harness":"DeepSeek Harness standard mode","variant":"reasoning_effort=100","version":"1.1","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash"},{"name":"Terminal-Bench","score":85.8,"metric":"pass@1","harness":"DeepSeek Harness standard mode","variant":"reasoning_effort=100","version":"2.1","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash"},{"name":"DeepSWE","score":67.6,"metric":"resolved","harness":"DeepSeek Harness PTC mode","variant":"reasoning_effort=100","version":"1.1","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash"},{"name":"Terminal-Bench","score":85.8,"metric":"pass@1","harness":"DeepSeek Harness PTC mode","variant":"reasoning_effort=100","version":"2.1","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash"}]},"deepseek/deepseek-v4-pro-0423":{"id":"deepseek/deepseek-v4-pro-0423","name":"DeepSeek V4 Pro 0423","description":"DeepSeek V4 Pro initial snapshot with million-token context and support for thinking and non-thinking modes","family":"deepseek-thinking","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-04-23","last_updated":"2026-04-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000}},"deepseek/deepseek-ocr-2":{"id":"deepseek/deepseek-ocr-2","name":"DeepSeek OCR 2","description":"High-accuracy OCR model for extracting text from documents, screenshots, receipts, and natural scenes","attachment":true,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2026-01-27","last_updated":"2026-01-27","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":8192,"output":8192}},"deepseek/deepseek-chat-v3.1":{"id":"deepseek/deepseek-chat-v3.1","name":"DeepSeek V3.1","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-21","last_updated":"2025-08-21","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":32768}},"deepseek/deepseek-v4.1-flash-1017":{"id":"deepseek/deepseek-v4.1-flash-1017","name":"DeepSeek V4.1 Flash 1017","description":"October 17 update of DeepSeek V4.1 Flash with improved reasoning, coding, agentic tasks, and visual understanding","family":"deepseek-flash","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2026-10-17","last_updated":"2026-10-17","modalities":{"input":["text","image"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"license":"MIT","weights":[{"label":"Hugging Face","url":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017"}],"benchmarks":[{"name":"Humanity's Last Exam","score":42.6,"variant":"without tools","dataset":"full set","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017"},{"name":"Humanity's Last Exam","score":44,"variant":"without tools","dataset":"text-only subset","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017"},{"name":"Terminal-Bench","score":42.3,"harness":"Claude Code","version":"4.0","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017"},{"name":"DeepSWE","score":77,"harness":"Mini-SWE-V2","version":"1.1","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017"},{"name":"ProgramBench","score":29.1,"source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017"},{"name":"NL2Repo","score":68.2,"source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017"},{"name":"AutomationBench","score":60.5,"version":"1.0.6","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017"},{"name":"Agents' Last Exam","score":35.4,"source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017"},{"name":"Apex Bench","score":72.1,"version":"1.1","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017"},{"name":"Chartography","score":83,"variant":"with tools","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017"},{"name":"BabyVision","score":92.5,"variant":"with tools","source":"https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017"},{"name":"ZeroBench","score":51,"metric":"pass@5","variant":"with tools","dataset":"main","source":"https://api-docs.deepseek.com/updates/"}]},"deepseek/deepseek-v3.2":{"id":"deepseek/deepseek-v3.2","name":"DeepSeek V3.2","description":"Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use","family":"deepseek","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2024-07","release_date":"2025-12-01","last_updated":"2025-12-01","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":64000},"license":"MIT License","weights":[{"label":"Hugging Face","url":"https://huggingface.co/deepseek-ai/DeepSeek-V3.2"}]},"deepseek/deepseek-r1-0528":{"id":"deepseek/deepseek-r1-0528","name":"DeepSeek R1 0528","description":"May 2025 DeepSeek R1 update with improved mathematics, coding, and reasoning","family":"deepseek-thinking","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2025-05-28","last_updated":"2025-05-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":163840,"output":32768},"license":"MIT","weights":[{"label":"Hugging Face","url":"https://huggingface.co/deepseek-ai/DeepSeek-R1-0528"}]},"deepseek/deepseek-chat":{"id":"deepseek/deepseek-chat","name":"DeepSeek Chat","description":"DeepSeek chat model for instruction following, coding, and analysis","family":"deepseek","attachment":false,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2025-09","release_date":"2025-12-01","last_updated":"2026-02-28","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1000000,"output":384000},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/deepseek-ai/DeepSeek-V3.2"}],"benchmarks":[{"name":"Aider Polyglot","score":70.2,"metric":"percent correct","source":"https://aider.chat/docs/leaderboards/","date":"2025-10-03"}]},"xiaomi/mimo-v2.5-pro-ultraspeed":{"id":"xiaomi/mimo-v2.5-pro-ultraspeed","name":"MiMo-V2.5-Pro-UltraSpeed","description":"MiMo pro model for strong multimodal reasoning and agent execution","family":"mimo","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-06-08","last_updated":"2026-06-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro-FP4-DFlash"}]},"xiaomi/mimo-v2-omni":{"id":"xiaomi/mimo-v2-omni","name":"MiMo-V2-Omni","description":"MiMo omni model for text, image, video, audio, and agents","family":"mimo","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text","image","audio","video","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":262144,"output":131072}},"xiaomi/mimo-v2.5-tts-voicedesign":{"id":"xiaomi/mimo-v2.5-tts-voicedesign","type":"speech","name":"MiMo-V2.5-TTS-VoiceDesign","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"mimo","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":8192,"output":8192}},"xiaomi/mimo-v2.5-tts-voiceclone":{"id":"xiaomi/mimo-v2.5-tts-voiceclone","type":"speech","name":"MiMo-V2.5-TTS-VoiceClone","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"mimo","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":8192,"output":8192}},"xiaomi/mimo-v2-tts":{"id":"xiaomi/mimo-v2-tts","type":"speech","name":"MiMo-V2-TTS","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"mimo","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":8192,"output":8192}},"xiaomi/mimo-v2.5-tts":{"id":"xiaomi/mimo-v2.5-tts","type":"speech","name":"MiMo-V2.5-TTS","description":"Speech generation model for controllable voice, narration, and audio delivery","family":"mimo","attachment":false,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["audio"]},"open_weights":false,"limit":{"context":8192,"output":8192}},"xiaomi/mimo-v2-pro":{"id":"xiaomi/mimo-v2-pro","name":"MiMo-V2-Pro","description":"Earlier MiMo Pro model for multimodal agents, reasoning, and code tasks","family":"mimo","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-03-18","last_updated":"2026-03-18","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":1048576,"output":131072}},"xiaomi/mimo-v2.6-flash":{"id":"xiaomi/mimo-v2.6-flash","name":"MiMo-V2.6-Flash","description":"MiMo Flash model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"}],"benchmarks":[{"name":"DeepSWE","score":67.9,"metric":"score","version":"1.1","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"ProgramBench","score":26,"metric":"score","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"MiMo Code Bench","score":61.2,"metric":"score","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"AutomationBench","score":52.3,"metric":"score","version":"1.0.6","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"Toolathlon-Verified","score":73.6,"metric":"score","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"Agents' Last Exam","score":27.6,"metric":"score","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"Terminal-Bench","score":28.8,"metric":"score","version":"4.0","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"Terminal-Bench","score":87.6,"metric":"score","version":"2.1","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"OSWorld-Verified","score":80.8,"metric":"score","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"JobBench","score":61.2,"metric":"score","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"CyberGym","score":95.1,"metric":"score","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"MiMo Cyber Bench","score":77.2,"metric":"score","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"ExploitGym","score":6,"metric":"score","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"ExploitBench","score":25.3,"metric":"score","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"SEC-Bench Pro","score":47.5,"metric":"score","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"MiMo VisualCoding","score":71.5,"metric":"score","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"}]},"xiaomi/mimo-v2.6-pro-ultraspeed":{"id":"xiaomi/mimo-v2.6-pro-ultraspeed","name":"MiMo-V2.6-Pro-UltraSpeed","description":"MiMo pro model for strong multimodal reasoning and agent execution","family":"mimo","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Pro-RL"}]},"xiaomi/mimo-v2.6-pro":{"id":"xiaomi/mimo-v2.6-pro","name":"MiMo-V2.6-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Pro-RL"}],"benchmarks":[{"name":"DeepSWE","score":71.9,"metric":"score","variant":"MiMo-V2.6 Pro comparison column","version":"1.1","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"ProgramBench","score":26.5,"metric":"score","variant":"MiMo-V2.6 Pro comparison column","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"MiMo Code Bench","score":63.2,"metric":"score","variant":"MiMo-V2.6 Pro comparison column","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"AutomationBench","score":53.1,"metric":"score","variant":"MiMo-V2.6 Pro comparison column","version":"1.0.6","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"Toolathlon-Verified","score":76.9,"metric":"score","variant":"MiMo-V2.6 Pro comparison column","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"GDPval-AA","score":1673,"metric":"Elo","variant":"MiMo-V2.6 Pro comparison column","version":"2.1","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"Agents' Last Exam","score":31.6,"metric":"score","variant":"MiMo-V2.6 Pro comparison column","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"Terminal-Bench","score":34.9,"metric":"score","variant":"MiMo-V2.6 Pro comparison column","version":"4.0","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"Terminal-Bench","score":89.9,"metric":"score","variant":"MiMo-V2.6 Pro comparison column","version":"2.1","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"OSWorld-Verified","score":82,"metric":"score","variant":"MiMo-V2.6 Pro comparison column","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"JobBench","score":62,"metric":"score","variant":"MiMo-V2.6 Pro comparison column","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"CyberGym","score":94,"metric":"score","variant":"MiMo-V2.6 Pro comparison column","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"MiMo Cyber Bench","score":80.2,"metric":"score","variant":"MiMo-V2.6 Pro comparison column","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"ExploitGym","score":17.8,"metric":"score","variant":"MiMo-V2.6 Pro comparison column","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"ExploitBench","score":47.9,"metric":"score","variant":"MiMo-V2.6 Pro comparison column","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"SEC-Bench Pro","score":66.3,"metric":"score","variant":"MiMo-V2.6 Pro comparison column","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"MiMo VisualCoding","score":72.3,"metric":"score","variant":"MiMo-V2.6 Pro comparison column","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"}]},"xiaomi/mimo-v2.5-pro":{"id":"xiaomi/mimo-v2.5-pro","name":"MiMo-V2.5-Pro","description":"Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution","family":"mimo","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro"}],"benchmarks":[{"name":"SWE-Bench Verified","score":78.9,"metric":"resolved","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro"},{"name":"SWE-Bench Pro","score":57.2,"metric":"resolve rate","source":"https://mimo.xiaomi.com/mimo-v2-5-pro/","date":"2026-04-22"},{"name":"GPQA Diamond","score":86.6,"metric":"accuracy","source":"https://mimo.xiaomi.com/mimo-v2-5-pro/","date":"2026-04-22"},{"name":"DeepSWE","score":19,"metric":"score","variant":"MiMo-V2.5 Pro comparison column","version":"1.1","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"ProgramBench","score":12.5,"metric":"score","variant":"MiMo-V2.5 Pro comparison column","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"MiMo Code Bench","score":40.4,"metric":"score","variant":"MiMo-V2.5 Pro comparison column","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"AutomationBench","score":16,"metric":"score","variant":"MiMo-V2.5 Pro comparison column","version":"1.0.6","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"Toolathlon-Verified","score":49.1,"metric":"score","variant":"MiMo-V2.5 Pro comparison column","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"GDPval-AA","score":1107,"metric":"Elo","variant":"MiMo-V2.5 Pro comparison column","version":"2.1","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"Agents' Last Exam","score":13.2,"metric":"score","variant":"MiMo-V2.5 Pro comparison column","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"Terminal-Bench","score":1.5,"metric":"score","variant":"MiMo-V2.5 Pro comparison column","version":"4.0","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"Terminal-Bench","score":65.2,"metric":"score","variant":"MiMo-V2.5 Pro comparison column","version":"2.1","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"JobBench","score":25,"metric":"score","variant":"MiMo-V2.5 Pro comparison column","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"CyberGym","score":40,"metric":"score","variant":"MiMo-V2.5 Pro comparison column","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"MiMo Cyber Bench","score":0,"metric":"score","variant":"MiMo-V2.5 Pro comparison column","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"ExploitGym","score":0.2,"metric":"score","variant":"MiMo-V2.5 Pro comparison column","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"ExploitBench","score":16.6,"metric":"score","variant":"MiMo-V2.5 Pro comparison column","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"},{"name":"SEC-Bench Pro","score":17.7,"metric":"score","variant":"MiMo-V2.5 Pro comparison column","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Flash-RL"}]},"xiaomi/mimo-v2-flash":{"id":"xiaomi/mimo-v2-flash","name":"MiMo-V2-Flash","description":"MiMo flash model for fast multimodal assistance and agent workflows","family":"mimo","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2024-12-01","release_date":"2025-12-16","last_updated":"2026-02-04","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":65536},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/XiaomiMiMo/MiMo-V2-Flash"}]},"xiaomi/mimo-v2.5":{"id":"xiaomi/mimo-v2.5","name":"MiMo-V2.5","description":"Open MiMo model for multimodal coding agents and long-context automation","family":"mimo","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2024-12","release_date":"2026-04-22","last_updated":"2026-04-22","modalities":{"input":["text","image","audio","video"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/XiaomiMiMo/MiMo-V2.5"}],"benchmarks":[{"name":"MiMo Coding Bench","score":71.8,"metric":"score","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.5"},{"name":"Claw-Eval","score":62.3,"metric":"score","variant":"text","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.5"},{"name":"Terminal-Bench","score":65.8,"metric":"score","version":"2.0","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.5"},{"name":"SWE-Bench Pro","score":56.1,"metric":"score","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.5"},{"name":"CharXiv Reasoning","score":81,"metric":"score","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.5"},{"name":"MMMU-Pro","score":77.9,"metric":"score","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.5"},{"name":"HR-Bench","score":88.5,"metric":"score","variant":"4K","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.5"},{"name":"OmniDocBench","score":87.2,"metric":"score","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.5"},{"name":"Claw-Eval","score":23.8,"metric":"score","variant":"multimodal","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.5"},{"name":"Video-MME","score":87.7,"metric":"score","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.5"},{"name":"DailyOmni","score":83.5,"metric":"score","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.5"},{"name":"VideoHolmes","score":64,"metric":"score","source":"https://huggingface.co/XiaomiMiMo/MiMo-V2.5"}]},"sdaia/allam-2-7b":{"id":"sdaia/allam-2-7b","name":"ALLaM-2-7b","description":"ALLaM-2-7b instruction tuned model by SDAIA","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"release_date":"2025-01-23","last_updated":"2025-01-23","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":4096,"output":4096},"weights":[{"label":"Hugging Face","url":"https://huggingface.co/ALLaM-AI/ALLaM-2.0-7B-Instruct"}]},"rekaai/reka-edge":{"id":"rekaai/reka-edge","name":"Reka Edge","description":"Multimodal model for analyzing text, images, documents, and rich media","family":"reka","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"release_date":"2026-03-11","last_updated":"2026-03-11","modalities":{"input":["image","text","video"],"output":["text"]},"open_weights":true,"limit":{"context":16384,"output":14745}},"rekaai/reka-flash-3":{"id":"rekaai/reka-flash-3","name":"Reka Flash 3","description":"Efficient model for low-latency assistance, extraction, and routine automation","family":"reka","attachment":false,"reasoning":true,"tool_call":false,"temperature":true,"knowledge":"2025-01-31","release_date":"2025-03-10","last_updated":"2025-03-10","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":65536,"output":58982}},"cognitivecomputations/dolphin-mistral-24b-venice-edition":{"id":"cognitivecomputations/dolphin-mistral-24b-venice-edition","name":"Uncensored","description":"Mistral model for multilingual chat, reasoning, and tool-assisted workflows","family":"mistral","attachment":false,"reasoning":false,"tool_call":false,"temperature":true,"knowledge":"2024-04-30","release_date":"2025-07-09","last_updated":"2025-07-09","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":128000,"output":8192}},"writer/palmyra-x4":{"id":"writer/palmyra-x4","name":"Palmyra X4","description":"Enterprise language model for workflow automation, coding, data analysis, and tool use","family":"palmyra","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2024-10-09","last_updated":"2025-04-28","modalities":{"input":["text"],"output":["text"]},"open_weights":false,"limit":{"context":128000,"output":4096}},"writer/palmyra-x5":{"id":"writer/palmyra-x5","name":"Palmyra X5","description":"Reasoning model for deliberate analysis, multi-step problem solving, and tool use","family":"palmyra","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2025-04-28","last_updated":"2025-04-28","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":8192}},"reflection/beam-1.0-501b-a23b":{"id":"reflection/beam-1.0-501b-a23b","name":"Beam 1.0 501B A23B","description":"Sparse mixture-of-experts reasoning model with 23B active parameters for coding and agentic workflows","family":"beam","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2026-06-30","release_date":"2026-10-05","last_updated":"2026-10-16","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":1048576,"output":131072},"license":"Apache-2.0","links":[{"label":"Model card","url":"https://huggingface.co/reflection/Beam-1.0-501B-A23B","type":"model_card"},{"label":"Launch announcement","url":"https://reflection.ai/blog/beam-is-available","type":"announcement"},{"label":"Introducing Beam","url":"https://reflection.ai/blog/introducing-beam","type":"blog"},{"label":"API docs","url":"https://developers.reflection.ai/models","type":"docs"}],"weights":[{"label":"Hugging Face","url":"https://huggingface.co/reflection/Beam-1.0-501B-A23B"},{"label":"Hugging Face (FP8)","url":"https://huggingface.co/reflection/Beam-1.0-501B-A23B-FP8","quantization":"fp8"},{"label":"Hugging Face (NVFP4)","url":"https://huggingface.co/reflection/Beam-1.0-501B-A23B-NVFP4","quantization":"nvfp4"}],"benchmarks":[{"name":"DeepSWE","score":43.3,"variant":"BF16","version":"1.1","source":"https://huggingface.co/reflection/Beam-1.0-501B-A23B","date":"2026-10-16"},{"name":"SWE-Bench Pro","score":76.5,"variant":"BF16","dataset":"hard","version":"2","source":"https://huggingface.co/reflection/Beam-1.0-501B-A23B","date":"2026-10-16"},{"name":"SWE-Bench Pro","score":66.1,"variant":"BF16","version":"1","source":"https://huggingface.co/reflection/Beam-1.0-501B-A23B","date":"2026-10-16"},{"name":"Terminal-Bench","score":80.1,"variant":"BF16","version":"2.1","source":"https://huggingface.co/reflection/Beam-1.0-501B-A23B","date":"2026-10-16"},{"name":"SWE-Atlas Codebase QnA","score":31.8,"variant":"BF16","source":"https://huggingface.co/reflection/Beam-1.0-501B-A23B","date":"2026-10-16"},{"name":"SWE-Bench Multilingual","score":78.3,"variant":"BF16","source":"https://huggingface.co/reflection/Beam-1.0-501B-A23B","date":"2026-10-16"},{"name":"SWE-Bench Verified","score":81.6,"variant":"BF16","source":"https://huggingface.co/reflection/Beam-1.0-501B-A23B","date":"2026-10-16"},{"name":"AIME 2026","score":97.7,"variant":"BF16","source":"https://huggingface.co/reflection/Beam-1.0-501B-A23B","date":"2026-10-16"},{"name":"Humanity's Last Exam","score":36,"variant":"BF16; no tools","source":"https://huggingface.co/reflection/Beam-1.0-501B-A23B","date":"2026-10-16"},{"name":"Humanity's Last Exam","score":51.1,"variant":"BF16; with tools","source":"https://huggingface.co/reflection/Beam-1.0-501B-A23B","date":"2026-10-16"},{"name":"SciCode","score":49.2,"variant":"BF16","source":"https://huggingface.co/reflection/Beam-1.0-501B-A23B","date":"2026-10-16"},{"name":"CritPt","score":16.3,"variant":"BF16","source":"https://huggingface.co/reflection/Beam-1.0-501B-A23B","date":"2026-10-16"},{"name":"GPQA Diamond","score":90.6,"variant":"BF16","source":"https://huggingface.co/reflection/Beam-1.0-501B-A23B","date":"2026-10-16"},{"name":"LiveCodeBench","score":88.3,"variant":"BF16","source":"https://huggingface.co/reflection/Beam-1.0-501B-A23B","date":"2026-10-16"},{"name":"AutomationBench","score":39.3,"variant":"BF16","dataset":"public","source":"https://huggingface.co/reflection/Beam-1.0-501B-A23B","date":"2026-10-16"},{"name":"MCP Atlas","score":80.8,"variant":"BF16","source":"https://huggingface.co/reflection/Beam-1.0-501B-A23B","date":"2026-10-16"},{"name":"τ3 Banking","score":36.1,"variant":"BF16","source":"https://huggingface.co/reflection/Beam-1.0-501B-A23B","date":"2026-10-16"},{"name":"BrowseComp","score":76.8,"variant":"BF16; with context management","source":"https://huggingface.co/reflection/Beam-1.0-501B-A23B","date":"2026-10-16"},{"name":"DeepSearchQA","score":80.5,"variant":"BF16; with context management","source":"https://huggingface.co/reflection/Beam-1.0-501B-A23B","date":"2026-10-16"},{"name":"AA-LCR","score":78.2,"variant":"BF16","source":"https://huggingface.co/reflection/Beam-1.0-501B-A23B","date":"2026-10-16"},{"name":"LongBench","score":67.5,"variant":"BF16","version":"2","source":"https://huggingface.co/reflection/Beam-1.0-501B-A23B","date":"2026-10-16"},{"name":"IFBench","score":79.6,"variant":"BF16","source":"https://huggingface.co/reflection/Beam-1.0-501B-A23B","date":"2026-10-16"},{"name":"AA-Omniscience","score":10.8,"metric":"index","variant":"BF16","dataset":"public split","source":"https://huggingface.co/reflection/Beam-1.0-501B-A23B","date":"2026-10-16"}]},"typesafe/jev-1.14.0":{"id":"typesafe/jev-1.14.0","type":"decision","name":"Jev 1.14","description":"System One model for fast, typed probabilistic decisions over text or structured state","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-10-14","last_updated":"2026-10-14","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":64000,"output":0}},"typesafe/jev-latest":{"id":"typesafe/jev-latest","type":"decision","name":"Jev","description":"System One model for fast, typed probabilistic decisions over text or structured state","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-15","last_updated":"2026-09-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":64000,"output":0}},"typesafe/jev-1.13.0":{"id":"typesafe/jev-1.13.0","type":"decision","name":"Jev 1.13","description":"System One model for fast, typed probabilistic decisions over text or structured state","attachment":true,"reasoning":false,"tool_call":false,"temperature":false,"release_date":"2026-09-15","last_updated":"2026-09-15","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":64000,"output":0}},"vivgrid/viv-fast":{"id":"vivgrid/viv-fast","name":"Viv Fast","description":"Fast coding model","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2026-09","release_date":"2026-09-21","last_updated":"2026-09-21","modalities":{"input":["text","image"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":256000},"links":[{"label":"Model docs","url":"https://docs.vivgrid.com/models/viv-fast"}]},"anthropic/claude-sonnet-4-20250514":{"id":"anthropic/claude-sonnet-4-20250514","name":"Claude Sonnet 4","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"benchmarks":[{"name":"Aider Polyglot","score":61.3,"metric":"percent correct","source":"https://aider.chat/docs/leaderboards/","date":"2025-05-24"}]},"anthropic/claude-opus-4-7":{"id":"anthropic/claude-opus-4-7","name":"Claude Opus 4.7","description":"Stronger Opus tier for advanced software work and high-stakes reasoning","family":"claude-opus","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-04-16","last_updated":"2026-04-16","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"benchmarks":[{"name":"SWE-Bench Pro","score":64.3,"metric":"resolve rate","source":"https://www.anthropic.com/news/claude-opus-4-8","date":"2026-05-28"},{"name":"Terminal-Bench","score":66.1,"metric":"success rate","harness":"Terminus-2","version":"2.1","source":"https://www.anthropic.com/news/claude-opus-4-8","date":"2026-05-28"},{"name":"SWE-Atlas Refactoring","score":48.57,"metric":"score","harness":"Claude Code","source":"https://labs.scale.com/leaderboard/sweatlas-refactoring"},{"name":"Artificial Analysis Coding Agent Index","score":66.6,"metric":"average pass@1","harness":"Claude Code","variant":"max","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"SWE-Atlas Codebase QnA","score":81,"metric":"pass@1","harness":"Claude Code","variant":"max","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"SWE-Bench Pro","score":44.9,"metric":"pass@1","harness":"Claude Code","variant":"max","dataset":"hard-aa","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"Terminal-Bench","score":73.8,"metric":"pass@1","harness":"Claude Code","variant":"max","version":"2.1","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"Artificial Analysis Coding Agent Index","score":61.2,"metric":"average pass@1","harness":"Cursor CLI","variant":"medium","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"SWE-Atlas Codebase QnA","score":78.4,"metric":"pass@1","harness":"Cursor CLI","variant":"medium","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"SWE-Bench Pro","score":34.4,"metric":"pass@1","harness":"Cursor CLI","variant":"medium","dataset":"hard-aa","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"Terminal-Bench","score":70.6,"metric":"pass@1","harness":"Cursor CLI","variant":"medium","version":"2.1","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"Artificial Analysis Coding Agent Index","score":59.9,"metric":"average pass@1","harness":"Claude Code","variant":"medium","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"SWE-Atlas Codebase QnA","score":71.7,"metric":"pass@1","harness":"Claude Code","variant":"medium","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"SWE-Bench Pro","score":36.4,"metric":"pass@1","harness":"Claude Code","variant":"medium","dataset":"hard-aa","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"Terminal-Bench","score":71.4,"metric":"pass@1","harness":"Claude Code","variant":"medium","version":"2.1","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"GPQA Diamond","score":94.2,"metric":"accuracy","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"Humanity's Last Exam","score":46.9,"metric":"accuracy","variant":"no tools","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"Humanity's Last Exam","score":54.7,"metric":"accuracy","variant":"with tools","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"},{"name":"OSWorld-Verified","score":78,"metric":"success rate","source":"https://openai.com/index/introducing-gpt-5-5/","date":"2026-04-23"}]},"anthropic/claude-opus-4-5-20251101":{"id":"anthropic/claude-opus-4-5-20251101","name":"Claude Opus 4.5","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"benchmarks":[{"name":"SWE-Bench Pro","score":45.89,"metric":"resolve rate","dataset":"public","source":"https://labs.scale.com/leaderboard/swe_bench_pro_public"}]},"anthropic/claude-sonnet-5-5":{"id":"anthropic/claude-sonnet-5-5","name":"Claude Sonnet 5.5","description":"Fast Claude model for everyday coding, agents, and knowledge work","family":"claude-sonnet","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-28","last_updated":"2026-09-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"benchmarks":[{"name":"Terminal-Bench","score":70.6,"metric":"score","version":"4.0","source":"https://www.anthropic.com/claude-sonnet-5-5","date":"2026-09-28"},{"name":"FrontierCode","score":46.2,"metric":"score","variant":"max effort","dataset":"Main","version":"1.1","source":"https://www.anthropic.com/claude-sonnet-5-5","date":"2026-09-28"},{"name":"CursorBench","score":55.5,"metric":"score","version":"4.0","source":"https://www.anthropic.com/claude-sonnet-5-5","date":"2026-09-28"},{"name":"GDPval-AA","score":1844,"metric":"Elo","version":"2.1","source":"https://www.anthropic.com/claude-sonnet-5-5","date":"2026-09-28"},{"name":"Humanity's Last Exam","score":64.5,"metric":"score","variant":"with tools","source":"https://www.anthropic.com/claude-sonnet-5-5","date":"2026-09-28"},{"name":"OSWorld","score":80.1,"metric":"partial score","version":"2.1","source":"https://www.anthropic.com/claude-sonnet-5-5","date":"2026-09-28"}]},"anthropic/claude-haiku-4-5":{"id":"anthropic/claude-haiku-4-5","name":"Claude Haiku 4.5 (latest)","description":"Fast Claude lane for lightweight agents, office tasks, and responsive chat","family":"claude-haiku","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"benchmarks":[{"name":"SWE-Bench Pro","score":39.45,"metric":"resolve rate","dataset":"public","source":"https://labs.scale.com/leaderboard/swe_bench_pro_public"}]},"anthropic/claude-opus-5":{"id":"anthropic/claude-opus-5","name":"Claude Opus 5","description":"Strongest Claude Opus model for coding, agents, and professional work","family":"claude-opus","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2026-05","release_date":"2026-07-24","last_updated":"2026-07-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"benchmarks":[{"name":"SWE-Bench Verified","score":96,"metric":"resolved","source":"https://www.anthropic.com/claude-opus-5-system-card","date":"2026-07-24"},{"name":"SWE-Bench Pro","score":79.2,"metric":"resolve rate","source":"https://www.anthropic.com/claude-opus-5-system-card","date":"2026-07-24"},{"name":"SWE-Bench Multilingual","score":89.5,"metric":"resolve rate","source":"https://www.anthropic.com/claude-opus-5-system-card","date":"2026-07-24"},{"name":"SWE-Bench Multimodal","score":59.4,"metric":"resolve rate","source":"https://www.anthropic.com/claude-opus-5-system-card","date":"2026-07-24"},{"name":"DeepSWE","score":68.8,"metric":"resolve rate","version":"1.1","source":"https://www.anthropic.com/claude-opus-5-system-card","date":"2026-07-24"},{"name":"FrontierCode","score":53.4,"metric":"mean@5","variant":"medium effort","dataset":"Main","version":"1.1","source":"https://www.anthropic.com/claude-opus-5-system-card","date":"2026-07-24"},{"name":"Frontier-Bench","score":43.3,"metric":"mean reward","harness":"mini-SWE-agent","variant":"max effort","version":"v0.1","source":"https://www.anthropic.com/claude-opus-5-system-card","date":"2026-07-24"},{"name":"BrowseComp","score":90.8,"metric":"accuracy","variant":"single agent","source":"https://www.anthropic.com/claude-opus-5-system-card","date":"2026-07-24"},{"name":"Humanity's Last Exam","score":56.3,"metric":"accuracy","variant":"no tools","source":"https://www.anthropic.com/claude-opus-5-system-card","date":"2026-07-24"},{"name":"Humanity's Last Exam","score":64.7,"metric":"accuracy","variant":"with tools","source":"https://www.anthropic.com/claude-opus-5-system-card","date":"2026-07-24"},{"name":"DeepSearchQA","score":95,"metric":"F1","variant":"max effort","source":"https://www.anthropic.com/news/claude-opus-5","date":"2026-07-24"},{"name":"OSWorld","score":70.6,"metric":"success rate","version":"2.0","source":"https://www.anthropic.com/claude-opus-5-system-card","date":"2026-07-24"},{"name":"GDPval-AA","score":1861,"metric":"Elo","variant":"max effort","version":"v2","source":"https://www.anthropic.com/claude-opus-5-system-card","date":"2026-07-24"},{"name":"AA-Briefcase","score":1720,"metric":"Elo","variant":"max effort","source":"https://www.anthropic.com/claude-opus-5-system-card","date":"2026-07-24"},{"name":"AutomationBench","score":26,"metric":"success rate","variant":"max effort","source":"https://www.anthropic.com/claude-opus-5-system-card","date":"2026-07-24"},{"name":"ARC-AGI-1","score":97.5,"metric":"accuracy","variant":"max effort","source":"https://www.anthropic.com/claude-opus-5-system-card","date":"2026-07-24"},{"name":"ARC-AGI-2","score":90.4,"metric":"accuracy","variant":"max effort","source":"https://www.anthropic.com/claude-opus-5-system-card","date":"2026-07-24"},{"name":"ARC-AGI-3","score":30.2,"metric":"RHAE","variant":"high effort","source":"https://www.anthropic.com/claude-opus-5-system-card","date":"2026-07-24"},{"name":"HealthBench Professional","score":59.8,"metric":"score","variant":"max effort","source":"https://www.anthropic.com/claude-opus-5-system-card","date":"2026-07-24"}]},"anthropic/claude-3-7-sonnet-20250219":{"id":"anthropic/claude-3-7-sonnet-20250219","name":"Claude Sonnet 3.7","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2024-10-31","release_date":"2025-02-19","last_updated":"2025-02-19","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"benchmarks":[{"name":"Aider Polyglot","score":64.9,"metric":"percent correct","source":"https://aider.chat/docs/leaderboards/","date":"2025-02-24"}]},"anthropic/claude-haiku-4-5-20251001":{"id":"anthropic/claude-haiku-4-5-20251001","name":"Claude Haiku 4.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-02-28","release_date":"2025-10-15","last_updated":"2025-10-15","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000}},"anthropic/claude-3-5-sonnet-20241022":{"id":"anthropic/claude-3-5-sonnet-20241022","name":"Claude Sonnet 3.5 v2","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-04-30","release_date":"2024-10-22","last_updated":"2024-10-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8192},"benchmarks":[{"name":"Aider Polyglot","score":51.6,"metric":"percent correct","source":"https://aider.chat/docs/leaderboards/","date":"2025-01-17"}]},"anthropic/claude-sonnet-4-6":{"id":"anthropic/claude-sonnet-4-6","name":"Claude Sonnet 4.6","description":"Claude workhorse for coding agents, careful analysis, and production cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-08-31","release_date":"2026-02-17","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":64000},"benchmarks":[{"name":"SWE-Atlas Codebase QnA","score":31.2,"metric":"score","harness":"Claude Code","source":"https://labs.scale.com/leaderboard/sweatlas-qna"},{"name":"SWE-Atlas Refactoring","score":32.21,"metric":"score","harness":"Claude Code","source":"https://labs.scale.com/leaderboard/sweatlas-refactoring"},{"name":"SWE-Atlas Test Writing","score":31.76,"metric":"score","harness":"Claude Code","source":"https://labs.scale.com/leaderboard/sweatlas-tw"},{"name":"Artificial Analysis Coding Agent Index","score":49.4,"metric":"average pass@1","harness":"Claude Code","variant":"medium","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"SWE-Atlas Codebase QnA","score":70.3,"metric":"pass@1","harness":"Claude Code","variant":"medium","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"SWE-Bench Pro","score":14.9,"metric":"pass@1","harness":"Claude Code","variant":"medium","dataset":"hard-aa","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"Terminal-Bench","score":63.1,"metric":"pass@1","harness":"Claude Code","variant":"medium","version":"2.1","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"Terminal-Bench","score":67,"metric":"success rate","harness":"Terminus-2","version":"2.1","source":"https://www.anthropic.com/news/claude-sonnet-5","date":"2026-06-30"},{"name":"Humanity's Last Exam","score":34.6,"metric":"accuracy","variant":"no tools","source":"https://www.anthropic.com/news/claude-sonnet-5","date":"2026-06-30"},{"name":"Humanity's Last Exam","score":46.8,"metric":"accuracy","variant":"with tools","source":"https://www.anthropic.com/news/claude-sonnet-5","date":"2026-06-30"},{"name":"OSWorld-Verified","score":78.5,"metric":"success rate","source":"https://www.anthropic.com/news/claude-sonnet-5","date":"2026-06-30"}]},"anthropic/claude-3-haiku-20240307":{"id":"anthropic/claude-3-haiku-20240307","name":"Claude Haiku 3","description":"Legacy model retained for compatibility with older integrations","family":"claude-haiku","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2023-08-31","release_date":"2024-03-13","last_updated":"2024-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":4096}},"anthropic/claude-opus-4-1-20250805":{"id":"anthropic/claude-opus-4-1-20250805","name":"Claude Opus 4.1","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000}},"anthropic/claude-opus-5-5":{"id":"anthropic/claude-opus-5-5","name":"Claude Opus 5.5","description":"Claude model for long-running agentic coding and knowledge work","family":"claude-opus","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-22","last_updated":"2026-09-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"benchmarks":[{"name":"Terminal-Bench","score":66.4,"metric":"score","variant":"xhigh effort; production safeguards with fallback","version":"4.0","source":"https://www.anthropic.com/claude-opus-5-5","date":"2026-09-22"},{"name":"FrontierCode","score":54.4,"metric":"score","variant":"max effort; production safeguards with fallback","dataset":"Main","version":"1.1","source":"https://www.anthropic.com/claude-opus-5-5","date":"2026-09-22"},{"name":"CursorBench","score":57.8,"metric":"score","variant":"max effort; production safeguards with fallback","version":"4.0","source":"https://www.anthropic.com/claude-opus-5-5","date":"2026-09-22"},{"name":"GDPval-AA","score":1846,"metric":"Elo","variant":"max effort; production safeguards with fallback","version":"2.1","source":"https://www.anthropic.com/claude-opus-5-5","date":"2026-09-22"},{"name":"AutomationBench","score":40,"metric":"pass rate","variant":"max effort; production safeguards; no fallback","source":"https://www.anthropic.com/claude-opus-5-5","date":"2026-09-22"},{"name":"Humanity's Last Exam","score":67.7,"metric":"score","variant":"max effort; with tools; production safeguards with fallback","source":"https://www.anthropic.com/claude-opus-5-5","date":"2026-09-22"},{"name":"Terminal-Bench-Science","score":58.7,"metric":"score","variant":"max effort; production safeguards with fallback","version":"0.1","source":"https://www.anthropic.com/claude-opus-5-5","date":"2026-09-22"},{"name":"OSWorld","score":81.8,"metric":"partial score","variant":"max effort; production safeguards with fallback","version":"2.0","source":"https://www.anthropic.com/claude-opus-5-5","date":"2026-09-22"},{"name":"Chartography","score":89,"metric":"score","variant":"max effort; with tools; production safeguards with fallback","source":"https://www.anthropic.com/claude-opus-5-5","date":"2026-09-22"}]},"anthropic/claude-opus-4-6":{"id":"anthropic/claude-opus-4-6","name":"Claude Opus 4.6","description":"High-end Claude for difficult coding, planning, and slower expert reasoning","family":"claude-opus","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-05-31","release_date":"2026-02-05","last_updated":"2026-03-13","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"benchmarks":[{"name":"SWE-Bench Pro","score":51.9,"metric":"resolve rate","dataset":"public","source":"https://labs.scale.com/leaderboard/swe_bench_pro_public"},{"name":"SWE-Atlas Codebase QnA","score":33.3,"metric":"score","harness":"Claude Code","source":"https://labs.scale.com/leaderboard/sweatlas-qna"},{"name":"SWE-Atlas Codebase QnA","score":30,"metric":"score","harness":"Mini-SWE-Agent","source":"https://labs.scale.com/leaderboard/sweatlas-qna"},{"name":"SWE-Atlas Refactoring","score":35.58,"metric":"score","harness":"Claude Code","source":"https://labs.scale.com/leaderboard/sweatlas-refactoring"},{"name":"SWE-Atlas Test Writing","score":36.67,"metric":"score","harness":"Claude Code","source":"https://labs.scale.com/leaderboard/sweatlas-tw"},{"name":"SWE-Atlas Test Writing","score":36.08,"metric":"score","harness":"Mini-SWE-Agent","source":"https://labs.scale.com/leaderboard/sweatlas-tw"},{"name":"Artificial Analysis Coding Agent Index","score":51.3,"metric":"average pass@1","harness":"Claude Code","variant":"medium","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"SWE-Atlas Codebase QnA","score":71.9,"metric":"pass@1","harness":"Claude Code","variant":"medium","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"SWE-Bench Pro","score":11.8,"metric":"pass@1","harness":"Claude Code","variant":"medium","dataset":"hard-aa","source":"https://artificialanalysis.ai/agents/coding-agents"},{"name":"Terminal-Bench","score":70.2,"metric":"pass@1","harness":"Claude Code","variant":"medium","version":"2.1","source":"https://artificialanalysis.ai/agents/coding-agents"}]},"anthropic/claude-sonnet-5":{"id":"anthropic/claude-sonnet-5","name":"Claude Sonnet 5","description":"Everyday Claude agent model for coding, planning, browsing, and general work","family":"claude-sonnet","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-30","last_updated":"2026-06-30","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"benchmarks":[{"name":"SWE-Bench Verified","score":85.2,"metric":"resolved","source":"https://www.anthropic.com/news/claude-sonnet-5","date":"2026-06-30"},{"name":"SWE-Bench Pro","score":63.2,"metric":"resolve rate","source":"https://www.anthropic.com/news/claude-sonnet-5","date":"2026-06-30"},{"name":"SWE-Bench Multilingual","score":78.3,"metric":"resolve rate","source":"https://www.anthropic.com/news/claude-sonnet-5","date":"2026-06-30"},{"name":"Terminal-Bench","score":80.4,"metric":"success rate","harness":"Terminus-2","version":"2.1","source":"https://www.anthropic.com/news/claude-sonnet-5","date":"2026-06-30"},{"name":"OSWorld-Verified","score":81.2,"metric":"success rate","source":"https://www.anthropic.com/news/claude-sonnet-5","date":"2026-06-30"},{"name":"BrowseComp","score":84.7,"metric":"accuracy","variant":"single agent","source":"https://www.anthropic.com/news/claude-sonnet-5","date":"2026-06-30"},{"name":"FrontierCode","score":38.8,"metric":"pass rate","version":"v1","source":"https://www.anthropic.com/news/claude-sonnet-5","date":"2026-06-30"}]},"anthropic/claude-opus-4-1":{"id":"anthropic/claude-opus-4-1","name":"Claude Opus 4.1 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-08-05","last_updated":"2025-08-05","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000}},"anthropic/claude-opus-4-20250514":{"id":"anthropic/claude-opus-4-20250514","name":"Claude Opus 4","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"benchmarks":[{"name":"Aider Polyglot","score":72,"metric":"percent correct","source":"https://aider.chat/docs/leaderboards/","date":"2025-05-25"}]},"anthropic/claude-mythos-5-1":{"id":"anthropic/claude-mythos-5-1","name":"Claude Mythos 5.1","description":"Restricted Claude model for advanced cybersecurity and biology research workflows","family":"claude-mythos","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000}},"anthropic/claude-sonnet-4-0":{"id":"anthropic/claude-sonnet-4-0","name":"Claude Sonnet 4 (latest)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"benchmarks":[{"name":"Aider Polyglot","score":61.3,"metric":"percent correct","source":"https://aider.chat/docs/leaderboards/","date":"2025-05-24"},{"name":"SWE-Bench Pro","score":42.7,"metric":"resolve rate","dataset":"public","source":"https://labs.scale.com/leaderboard/swe_bench_pro_public"}]},"anthropic/claude-mythos-5":{"id":"anthropic/claude-mythos-5","name":"Claude Mythos 5","description":"Restricted Claude model for advanced cybersecurity and biology research workflows","family":"claude-mythos","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000}},"anthropic/claude-haiku-5-5":{"id":"anthropic/claude-haiku-5-5","name":"Claude Haiku 5.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-10-07","last_updated":"2026-10-07","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"benchmarks":[{"name":"GDPval-AA","score":1620,"metric":"Elo","version":"2.1","source":"https://www.anthropic.com/claude-haiku-5-5","date":"2026-10-07"},{"name":"AA-Briefcase","score":1578,"metric":"Elo","version":"1.1","source":"https://www.anthropic.com/claude-haiku-5-5","date":"2026-10-07"},{"name":"OSWorld","score":72.4,"metric":"partial score","variant":"offline subset","version":"2.1","source":"https://www.anthropic.com/claude-haiku-5-5","date":"2026-10-07"},{"name":"Humanity's Last Exam","score":45.9,"metric":"score","variant":"no tools","source":"https://www.anthropic.com/claude-haiku-5-5","date":"2026-10-07"},{"name":"Humanity's Last Exam","score":57.4,"metric":"score","variant":"with tools","source":"https://www.anthropic.com/claude-haiku-5-5","date":"2026-10-07"},{"name":"Terminal-Bench","score":39.2,"metric":"score","version":"4.0","source":"https://www.anthropic.com/claude-haiku-5-5","date":"2026-10-07"},{"name":"FrontierCode","score":46.4,"metric":"score","dataset":"Main","version":"1.1","source":"https://www.anthropic.com/claude-haiku-5-5","date":"2026-10-07"},{"name":"Chartography","score":46.4,"metric":"score","variant":"no tools","source":"https://www.anthropic.com/claude-haiku-5-5","date":"2026-10-07"}]},"anthropic/claude-fable-5-1":{"id":"anthropic/claude-fable-5-1","name":"Claude Fable 5.1","description":"Claude model for demanding reasoning and long-horizon agentic work","family":"claude-fable","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2026-06","release_date":"2026-09-01","last_updated":"2026-09-01","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"benchmarks":[{"name":"Terminal-Bench-Science","score":52.6,"metric":"score","variant":"production safeguards with fallback","version":"0.1","source":"https://www.anthropic.com/claude-fable-and-mythos-5-1","date":"2026-09-01"},{"name":"Terminal-Bench","score":55.8,"metric":"score","variant":"production safeguards with fallback","version":"4.0","source":"https://www.anthropic.com/claude-fable-and-mythos-5-1","date":"2026-09-01"},{"name":"GDPval-AA","score":1853,"metric":"Elo","variant":"production safeguards with fallback","version":"2","source":"https://www.anthropic.com/claude-fable-and-mythos-5-1","date":"2026-09-01"},{"name":"OSWorld","score":77.9,"metric":"partial score","variant":"production safeguards; interventions score zero; no fallback","dataset":"August 2026 task release","version":"2.0","source":"https://www.anthropic.com/claude-fable-and-mythos-5-1","date":"2026-09-01"},{"name":"OSWorld","score":41.7,"metric":"strict success rate","variant":"production safeguards; interventions score zero; no fallback","dataset":"August 2026 task release","version":"2.0","source":"https://www.anthropic.com/claude-fable-and-mythos-5-1","date":"2026-09-01"},{"name":"Humanity's Last Exam","score":60.9,"metric":"score","variant":"without tools; production safeguards with fallback","source":"https://www.anthropic.com/claude-fable-and-mythos-5-1","date":"2026-09-01"},{"name":"Humanity's Last Exam","score":65,"metric":"score","variant":"with tools; production safeguards with fallback","source":"https://www.anthropic.com/claude-fable-and-mythos-5-1","date":"2026-09-01"},{"name":"AutomationBench","score":31.4,"metric":"pass rate","variant":"production safeguards with fallback","source":"https://www.anthropic.com/claude-fable-and-mythos-5-1","date":"2026-09-01"},{"name":"CursorBench","score":73.4,"metric":"score","variant":"max effort; production safeguards with fallback","version":"3.2.0","source":"https://www.anthropic.com/claude-fable-and-mythos-5-1","date":"2026-09-01"}]},"anthropic/claude-3-5-haiku-20241022":{"id":"anthropic/claude-3-5-haiku-20241022","name":"Claude Haiku 3.5","description":"Fast Claude model for responsive assistance, classification, and lightweight agents","family":"claude-haiku","attachment":true,"reasoning":false,"tool_call":true,"temperature":true,"knowledge":"2024-07-31","release_date":"2024-10-22","last_updated":"2024-10-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":8192},"benchmarks":[{"name":"Aider Polyglot","score":28,"metric":"percent correct","source":"https://aider.chat/docs/leaderboards/","date":"2024-12-21"}]},"anthropic/claude-fable-5":{"id":"anthropic/claude-fable-5","name":"Claude Fable 5","description":"Claude model for creative writing, analysis, and controlled agent workflows","family":"claude-fable","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2026-01-31","release_date":"2026-06-09","last_updated":"2026-06-09","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"benchmarks":[{"name":"SWE-Bench Pro","score":80.3,"metric":"resolve rate","source":"https://www.anthropic.com/news/claude-fable-5-mythos-5","date":"2026-06-09"},{"name":"SWE-Bench Verified","score":95,"metric":"resolved","source":"https://benchlm.ai/benchmarks/sweVerified"},{"name":"Terminal-Bench","score":88,"metric":"success rate","version":"2.1","source":"https://www.anthropic.com/news/claude-fable-5-mythos-5","date":"2026-06-09"},{"name":"Humanity's Last Exam","score":59,"metric":"accuracy","variant":"no tools","source":"https://www.anthropic.com/news/claude-fable-5-mythos-5","date":"2026-06-09"},{"name":"Humanity's Last Exam","score":64.5,"metric":"accuracy","variant":"with tools","source":"https://www.anthropic.com/news/claude-fable-5-mythos-5","date":"2026-06-09"},{"name":"OSWorld-Verified","score":85,"metric":"success rate","source":"https://www.anthropic.com/news/claude-fable-5-mythos-5","date":"2026-06-09"},{"name":"FrontierCode","score":29.3,"metric":"pass rate","variant":"high effort","dataset":"Diamond","source":"https://www.anthropic.com/news/claude-fable-5-mythos-5","date":"2026-06-09"},{"name":"GDPval-AA","score":1932,"metric":"Elo","source":"https://www.anthropic.com/news/claude-fable-5-mythos-5","date":"2026-06-09"},{"name":"AutomationBench","score":17.4,"metric":"success rate","source":"https://www.anthropic.com/news/claude-fable-5-mythos-5","date":"2026-06-09"}]},"anthropic/claude-opus-4-8":{"id":"anthropic/claude-opus-4-8","name":"Claude Opus 4.8","description":"Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"tool_call":true,"temperature":false,"knowledge":"2026-01","release_date":"2026-05-28","last_updated":"2026-05-28","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":1000000,"output":128000},"benchmarks":[{"name":"SWE-Bench Pro","score":69.2,"metric":"resolve rate","source":"https://www.anthropic.com/news/claude-opus-4-8","date":"2026-05-28"},{"name":"Terminal-Bench","score":74.6,"metric":"success rate","harness":"Terminus-2","version":"2.1","source":"https://www.anthropic.com/news/claude-opus-4-8","date":"2026-05-28"},{"name":"SWE-Bench Verified","score":88.6,"metric":"resolved","source":"https://benchlm.ai/benchmarks/sweVerified"},{"name":"Humanity's Last Exam","score":49.8,"metric":"accuracy","variant":"no tools","source":"https://www.anthropic.com/news/claude-fable-5-mythos-5","date":"2026-06-09"},{"name":"Humanity's Last Exam","score":57.9,"metric":"accuracy","variant":"with tools","source":"https://www.anthropic.com/news/claude-fable-5-mythos-5","date":"2026-06-09"},{"name":"OSWorld-Verified","score":83.4,"metric":"success rate","source":"https://www.anthropic.com/news/claude-fable-5-mythos-5","date":"2026-06-09"},{"name":"FrontierCode","score":13.4,"metric":"pass rate","variant":"high effort","dataset":"Diamond","source":"https://www.anthropic.com/news/claude-fable-5-mythos-5","date":"2026-06-09"}]},"anthropic/claude-opus-4-0":{"id":"anthropic/claude-opus-4-0","name":"Claude Opus 4 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-03-31","release_date":"2025-05-22","last_updated":"2025-05-22","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":32000},"benchmarks":[{"name":"Aider Polyglot","score":72,"metric":"percent correct","source":"https://aider.chat/docs/leaderboards/","date":"2025-05-25"}]},"anthropic/claude-opus-4-5":{"id":"anthropic/claude-opus-4-5","name":"Claude Opus 4.5 (latest)","description":"Flagship Claude model for deep reasoning, coding, and long-horizon agents","family":"claude-opus","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-05","release_date":"2025-11-24","last_updated":"2025-11-24","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000}},"anthropic/claude-sonnet-4-5":{"id":"anthropic/claude-sonnet-4-5","name":"Claude Sonnet 4.5 (latest)","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000},"benchmarks":[{"name":"SWE-Bench Pro","score":43.6,"metric":"resolve rate","dataset":"public","source":"https://labs.scale.com/leaderboard/swe_bench_pro_public"}]},"anthropic/claude-sonnet-4-5-20250929":{"id":"anthropic/claude-sonnet-4-5-20250929","name":"Claude Sonnet 4.5","description":"Balanced Claude model for coding, analysis, agent workflows, and cost control","family":"claude-sonnet","attachment":true,"reasoning":true,"tool_call":true,"temperature":true,"knowledge":"2025-07-31","release_date":"2025-09-29","last_updated":"2025-09-29","modalities":{"input":["text","image","pdf"],"output":["text"]},"open_weights":false,"limit":{"context":200000,"output":64000}},"motif-technologies/motif-3":{"id":"motif-technologies/motif-3","name":"Motif 3","description":"Motif 3 is a large-scale, decoder-only Mixture-of-Experts (MoE) language model with 314 billion total parameters and 13.2 billion parameters activated per token.","family":"motif","attachment":false,"reasoning":true,"tool_call":true,"temperature":true,"release_date":"2026-08-07","last_updated":"2026-08-07","modalities":{"input":["text"],"output":["text"]},"open_weights":true,"limit":{"context":262144,"output":262144},"license":"MIT","weights":[{"label":"Model weights","url":"https://huggingface.co/Motif-Technologies/Motif-3","format":"safetensors"}]}}