Mistral Large 4

Mistral Large 4 is the flagship model release from Mistral, announced as a public preview on the company’s blog. It is described as a 1 trillion-parameter natively multimodal model with 52 billion active parameters, making it Mistral’s largest and most capable model to date. The release continues Mistral’s focus on open-weight models and enterprise-grade AI platforms.

The core appeal is frontier performance across coding, agentic workflows, and multimodal tasks, with the company noting that the model is still improving as it is refined. Developers can try the preview through the Mistral Studio API today, while model weights are scheduled to drop at the end of the month, giving teams both hosted and self-hosted paths.

It is aimed at enterprises and developers who need strong reasoning and coding capabilities, especially those interested in sovereign or self-managed AI deployments. Organizations evaluating agentic assistants, multimodal applications, or coding copilots are the primary audience.

For practical use, start with the preview API in Mistral Studio to benchmark the model against your own workloads before weights become available. Keep in mind that this is a preview release, so behavior and performance may change, and production planning should account for the upcoming weight release and any licensing terms.

Overall, Mistral Large 4 is a notable step for open-weight frontier models, worth watching for teams that prioritize capability and deployment flexibility.