Mistral Medium 3 was offered with a path from hosted API use to customized deployment inside an organization's infrastructure. Mistral AI's May 2025 announcement describes hybrid, in-VPC and on-premises deployment, together with services for further training the model on company-specific material.
Those options address two separate enterprise decisions: where the model runs and how its behavior is adapted to a particular business. The announcement directs customers to the company for deployment and customization, so it should not be read as a release of freely downloadable weights.
Terms in this piece · Glossary
benchmark — A standard public test set for comparing AI models — the shared scoreboards behind every "model X beats model Y" claim.
