Mistral AI Releases Open Weights for Mistral Large 4, Empowering Sovereign Control of ‘Le Chonk’ Frontier Model

Mistral AI has made the open weights for its flagship model, Mistral Large 4 (ML4), available to the public, marking a significant step toward transparent and customizable AI. Nicknamed ‘Le Chonk,’ ML4 is the Paris-based company’s most capable large language model to date, featuring 1 trillion total parameters with 52 billion active parameters, and native multimodal capabilities. The release comes just three weeks after the model’s initial preview, underscoring Mistral’s rapid iteration pace. By open-sourcing the weights, Mistral enables enterprises and developers to run, fine-tune, and deploy the model on their own infrastructure, ensuring data sovereignty and reducing reliance on closed APIs. This move directly addresses growing concerns over vendor lock-in and data privacy, particularly for European organizations subject to strict regulations like GDPR. The ‘Le Chonk’ moniker reflects both the model’s massive scale and the company’s playful culture. The decision to release open weights positions Mistral as a strong alternative to proprietary models from OpenAI, Google, and Anthropic, while also competing with other open-weight efforts like Meta’s Llama series. For businesses seeking frontier-level performance with full control, ML4 offers a compelling option without sacrificing capability. As the AI landscape shifts toward greater transparency, Mistral’s move could accelerate adoption of open models across industries. Expect further refinements and community-driven innovations as developers get their hands on Le Chonk.