Mistral AI and NVIDIA announced a partnership on October 21, 2026, to accelerate engineering simulations. The collaboration integrates Mistral’s models with NVIDIA’s computing stack to improve workflows in computer-aided engineering and electronic design automation.
The initiative utilizes the Mistral Large 4 model, which features approximately 1 trillion parameters. Training for this effort involves around 3,800 NVIDIA Grace Blackwell GPUs. The companies also plan to release public benchmarks and datasets for physics and agentic AI.
NVIDIA’s Omniverse platform will be integrated to enhance 3D simulations and support SimReady assets. This partnership builds on previous collaborations, including the development of the Mistral NeMo model and Mistral’s membership in the NVIDIA Nemotron Coalition.
Key clients identified for these technological advancements include Airbus, BMW, and ASML. These companies represent the aerospace, automotive, and semiconductor manufacturing sectors, which are the primary targets for the new simulation tools.
