Mistral previews Large 4, a 1-trillion-parameter open-weight model it says leads US and European open models
Mistral launched a public preview of Mistral Large 4 on October 6, available through its preview API on Mistral Studio, with the model weights due by the end of the month. It is a natively multimodal model with 1 trillion total parameters and 49 billion active, trained from scratch on 3,800 NVIDIA Grace Blackwell GPUs in Mistral's own European datacentres. Mistral says it is competitive with the strongest open models globally and significantly outperforms any open-weight model developed in the US or Europe, with particular strength in coding, agentic workflows, cybersecurity, finance and law. Mistral says it worked with enterprises in manufacturing, engineering, logistics and other industries to train it, and that it will be able to run on private cloud or on-premise.
What it means for manufacturers and fabricators
Once the weights ship, a near-frontier model you can run on your own servers gets another strong, non-Chinese option, which matters if you want AI on your drawings and quotes without sending them to someone else's cloud.