Mistral releases open weights for Large 4 (Le Chonk) targeting cyber and coding

Mistral publicly released the open weights of its 1-trillion-parameter multimodal model Large 4 (Le Chonk) on 29 October 2026 after additional reinforcement learning. The model targets agentic coding and cybersecurity defense, with reported gains on Terminal-Bench and efficiency advantages over leading Chinese open-weight models such as GLM-5.3. Open weights enable self-hosted deployment, reducing dependence on third-party APIs for sensitive workloads.