Mistral releases weights for its trillion-parameter AI model, targeting cybersecurity

Mistral, the French AI company, has released the weights for its trillion-parameter open-source AI model, Mistral Large 4 (ML4), nicknamed "Le Chonk," allowing businesses and governments to download and deploy the system themselves. The move comes amid increasing competition between open-weight developers like Mistral and closed-model providers such as OpenAI and Anthropic, with Mistral arguing that organizations should not have to rely on external AI companies for advanced capabilities, especially in cybersecurity where control over data and systems is critical. Mistral claims ML4 achieves state-of-the-art performance at a fraction of the compute and capital of closed labs, and that it is comparable to Chinese open-weight competitor GLM-5.3 while using 47 percent fewer tokens per task. The model also reportedly outranks all other open models on Artificial Analysis' Coding Agent Index for agentic coding tasks. According to Mistral, ML4 brings sovereign cyber defense capabilities to enterprises and governments, enabling them to detect, analyze, and respond to threats across code and IT infrastructure. The weights are available on Hugging Face and via API. Pierre Stock, VP of Science at Mistral, emphasized that open-weighting ensures access to frontier intelligence cannot be cut off, and that organizations can deploy the model on-premises for real-time incident response. He also addressed concerns about rogue AI agents, acknowledging risks but cautioning against fearmongering and expressing confidence that the problem is solvable.

(Source：TechCrunch)