Mistral Drops a 1-Trillion-Parameter Bomb

🔥 What happened Mistral launched the public preview of Mistral Large 4, nicknamed "le Chonk." It's a 1-trillion-parameter natively multimodal model with 49B active parameters, trained from scratch on 3,800 Grace Blackwell GPUs in Mistral's own European datacenters. Weights drop by end of month. 💡 Why it matters Mistral claims ML4 beats every open US and EU model, and it scores 82% on a real-world vulnerability reproduction test — where Claude Opus 5.5 and GPT-6 Astra score near zero because they refuse the task. Open weights plus self-deployment could be the fix security teams have been waiting for. ⚡ Our take A European open-weight model out-cybering closed frontier models because those are too polite is the most interesting frontier story this month. Now Mistral just has to actually ship the weights.