Built for Coders,Trusted by Defenders.

Mistral has released the weights for Mistral Large 4 (Le Chonk), a 1-trillion-parameter, open-weight multimodal model. The model demonstrates leading performance in coding and defensive cybersecurity, outperforming other open-weight models on several key benchmarks. It introduces three distinct reasoning modes to give developers control over the trade-off between performance and inference cost. As a European-trained model, it is positioned as a trusted foundation for sovereign AI, allowing organizations to own, customize, and control their AI systems.