Mistral Large 4 replaces Large 3 as Mistral AI’s proprietary flagship, roughly eight months after that model shipped in December 2025. It’s a 190 billion parameter dense model with a 256,000 token context window, priced at $2 per million input tokens and $8 per million output tokens through Mistral’s own La Plateforme as well as Azure AI Foundry and AWS Bedrock. Mistral says the gains over Large 3 are concentrated in coding and multi-step reasoning, an area where the company has been trying to close the gap with GPT-5.6 and Claude Sonnet 5 rather than compete on raw scale.
Unlike the smaller open-weight Mistral models the company still ships regularly, Large 4 is proprietary and API-only, continuing the tiered strategy Mistral settled into with Large 2: open weights build the developer base, the top model funds the lab. European enterprises remain the core customer base, with Mistral leaning on EU data residency as a selling point against the mostly US-based frontier labs.