Mistral Unveils 'Le Chonk' AI Model It Says Rivals Best Open Models From China
Mistral has previewed Mistral Large 4, a 1-trillion-parameter open-weight model it calls the strongest developed outside China, trained on 4,000 Nvidia chips in Europe.
The model, nicknamed "Le Chonk," is entering preview for developers, cybersecurity teams and governments ahead of a broader release later this month. Mistral said ML4 was trained over two months on 4,000 Nvidia Grace Blackwell GPUs deployed in the company's own data centers in Europe.
In a statement, Mistral said that once the model's core parameters are released, ML4 will rank among the top open-weight models globally on aggregate benchmark performance. The company described it as the strongest open-weight model developed outside China by "a substantial margin." It also acknowledged that the model still lags behind the frontier in areas such as coding.
Guillaume Lample, Mistral's co-founder and chief scientist, said the model's capabilities would improve further as the company scales up its training capacity following its Series D fundraise. He said the cyber defense capabilities would allow enterprises and governments to defend themselves against threat actors that jailbreak closed models to carry out cyberattacks.
Slashdot's report on the release also directed readers to earlier comments by Mistral's chief executive, who said the debate over AI safety in the United States masks competitors' "negligence."