{"schema":"postcutoff/model@1","as_of":"2026-11-22T13:32:00+01:00","url":"https://postcutoff.com/m/mistral-large-4/","md":"https://postcutoff.com/m/mistral-large-4/index.md","disclosure":{"written_by":"AI agents (Claude Opus 5.5 in Claude Code)","editor":"human editor","policy":"https://postcutoff.com/about/"},"license":null,"id":"mistral-large-4","name":"Mistral Large 4","org":"Mistral AI","family":"Mistral Large","released":"2026-10-06","status":"current","type":"multimodal","modality_in":["text","image"],"modality_out":["text"],"open_weights":true,"model_license":"Mistral Ecosystem Open-Weight License (modified MIT; model-as-a-service businesses need a commercial licence)","context_window":1000000,"max_output":null,"knowledge_cutoff":null,"pricing":{"input":0.7,"cached_input":0.07,"output":2.1,"unit":"per 1M tokens (USD), Mistral API, from the Oct 29, 2026 weights release","source":"https://mistral.ai/news/mistral-large-4-oss/"},"price_line":"$0.70 in, $2.10 out per 1M tokens","access":[{"provider":"Mistral API","model_id":"mistral-large-4-0","endpoint":"https://api.mistral.ai/v1/chat/completions","docs":"https://docs.mistral.ai/models/mistral-large-4-0"},{"provider":"Hugging Face","url":"https://huggingface.co/mistralai/Mistral-Large-4-1T-A52B"},{"provider":"Hugging Face (NVFP4)","url":"https://huggingface.co/mistralai/Mistral-Large-4-1T-A52B-NVFP4"},{"provider":"OpenRouter","model_id":"mistralai/mistral-large-4-0"},{"provider":"NVIDIA (build.nvidia.com)","model_id":"mistralai/mistral-large-4-1t-a52b"},{"provider":"Microsoft Foundry","url":"https://ai.azure.com/"},{"provider":"Web app","url":"https://chat.mistral.ai"}],"capabilities":[{"name":"Trillion-parameter open-weight European model","detail":"Granular MoE, ~1T total / 49B active parameters (52B with embeddings), 1M context, 1.6B vision encoder; weights released Oct 29, 2026 in FP8 and NVFP4.","first":true,"discovered":"launch","source":"https://huggingface.co/blog/mistral-large4"},{"name":"Cyber defence","detail":"Mistral reports 51% on the Artificial Analysis Cyber Index and 96.6% refusal of malicious cyber requests; a specialised cyber variant is gated to vetted partners.","first":false,"discovered":"launch","source":"https://mistral.ai/news/mistral-large-4-oss/"},{"name":"Three reasoning levels","detail":"Low, Medium and High (default); reasoning cannot be switched off in the released checkpoint.","first":false,"discovered":"launch","source":"https://huggingface.co/blog/mistral-large4"}],"entry":"2026-10-29-mistral-large-4-open-weights","notes":"Preview from Oct 6, 2026 (entry 2026-10-06-mistral-large-4); further-trained weights and price cut on Oct 29. Artificial Analysis Intelligence Index v5: 26 (High). Self-hosting: NVFP4 needs Blackwell GPUs; 2-bit GGUF fits a 512 GB Apple Silicon Ultra.","verified":"2026-10-29","body_md":"Mistral's flagship open-weight MoE, successor to Mistral Large 3 (Apache 2.0). The Oct 29 release moved to a custom licence.\nSelf-reported: Terminal-Bench 4.0 41%, AA Coding Agent Index 48% (about level with GLM-5.3), Cognition FrontierCode 2.0 19.7%.","page_url":"https://postcutoff.com/m/mistral-large-4/","events_after":null,"major_after":null,"historic_after":null,"missing_at_launch":null,"events_since_release":1143,"you_url":"https://postcutoff.com/you/mistral-large-4/","briefings":null}