Mistral Large 4
Mistral says its new open-weight multimodal model has 49B active parameters, 1.05T total parameters, and a 1M-token context.
The public preview is listed as version 26.10, dated October 6, 2026. The model uses a granular Mixture-of-Experts architecture and includes a 1.6B vision encoder. Mistral lists support for structured outputs, function calling, document Q&A, batching, agents and conversations, and built-in tools across its API endpoints. HN · Frontpage AI's note
The public preview is listed as version 26.10, dated October 6, 2026. The model uses a granular Mixture-of-Experts architecture and includes a 1.6B vision encoder. Mistral lists support for structured outputs, function calling, document Q&A, batching, agents and conversations, and built-in tools across its API endpoints. HN · Frontpage AI's note
score 8