Mistral Large 4 packs 1.05 trillion parameters into an open-weight model
Mistral Large 4 is a state-of-the-art, open-weight, general-purpose multimodal model built on a granular Mixture-of-Experts architecture. It activates 49B parameters out of 1.05T total, includes a 1.6B vision encoder, and supports a 1M-token context window. Pricing starts at $0.68 per million input tokens and $2.09 per million output tokens, with cached input at $0.07. It ships with structured outputs, function calling, document QnA, and built-in tools.
Mistral Large 4 is a state-of-the-art, open-weight, general-purpose multimodal model with a granular Mixture-of-Experts architecture.