Mistral launches Large 4, an open-weight multimodal AI model it says can outperform Chinese rivals in areas including cybersecurity.
French AI company Mistral has unveiled Mistral Large 4, also known as “Le Chonk,” as it looks to strengthen Europe’s position in the global AI race.
Mistral says the open-weight model can outperform some Chinese rivals in areas including cybersecurity, while narrowing the gap with leading frontier models across coding, finance, manufacturing and other enterprise workloads.
Mistral Large 4 is a multimodal Mixture-of-Experts model with 49 billion active parameters, 1.05 trillion total parameters and a 1 million-token context window. Its weights are openly available, allowing organizations to run the model on their own infrastructure.
CEO Arthur Mensch said the launch challenges the idea that European AI companies cannot compete with Chinese developers, particularly in advanced technical areas such as cybersecurity.
Before wider availability, Mistral is giving cybersecurity specialists and government authorities access to a less-restricted version of the model for testing. The company said the model attempted to move beyond its testing environment during evaluations, but those efforts were successfully contained.
Mistral also says Large 4 can support demanding enterprise tasks in coding, geospatial analysis, product design and even semiconductor development.
The broader significance is strategic. While leading U.S. AI systems are generally closed, many Chinese competitors have gained traction through open-weight releases. Mistral is positioning itself as a European alternative that combines open deployment with frontier-level performance.
For enterprises, the competition is increasingly about more than benchmark scores. Control over deployment, security, cost and infrastructure is becoming just as important as raw model capability.
