As tensions rise over who has access to high-end artificial intelligence, French company Mistral has launched a new freely available model that it claims can compete with the best from the United States and China.
The new 1 billion parameter model, Mistral Large 4, nicknamed Le Chonk, can be used and customized by anyone. It is currently available in preview, and there will be a final version at the end of the month. Although Le Chonk is designed to compete with leading general-purpose models, it is optimized specifically for coding and cyber defense, as well as for specific tasks in manufacturing, finance, electrical engineering, and other niches.
“There are many areas that the other labs won’t focus on as much,” Guillaume Lample, co-founder and chief scientist at Mistral, tells WIRED. “There are so many domains where the models can be improved.”
Mistral presents the Le Chonk as by far the most capable open weight model developed outside of China, and “very, very close” to some proprietary models. While the US government has accused Chinese labs of abusing distillation (training a smaller model with the results of a larger one) to close the performance gap with OpenAI and Anthropic, Mistral claims to have trained its model from scratch.
It is already considerably cheaper for companies to run open models, which cost only the same as the compute they consume. By closing the performance gap in leading proprietary models and offering a competitive alternative to China releases, Mistral says, Le Chonk will eliminate the few remaining reasons a company might hesitate to go open source.
“Mistral is still in the race for the best model,” says Lample. “This is the main message.”
With less capital and fewer compute resources than OpenAI and Anthropic, Mistral has generally lagged behind in model performance, revenue, and release frequency. While U.S. labs charge a premium for access to their proprietary closed-weight models, Mistral generates revenue by charging pay-as-you-go fees for running models through its cloud and deploying engineers to help customers fine-tune the models to their specific needs. However, the French laboratory has recently been on a good run: in September, it raised a $3.3 billion financing round at a valuation of $24 billion, the largest fundraising ever by a European technology company. His earnings have reportedly increased 20-fold in the last year.
