Mistral and NVIDIA Partner to Develop AI-Native Engineering Workflows and Public Benchmarks

• Mistral and NVIDIA are collaborating to create AI-native engineering workflows by integrating Mistral Large 4, data generation, and physics modeling pipelines with NVIDIA CUDA-X libraries and NVIDIA Warp.
• The partnership will introduce public benchmarks and open datasets to standardize evaluations for physics models and agentic AI in computer-aided engineering (CAE) and electronic design automation (EDA).
• Mistral Large 4, known as Le Chonk, is a multimodal model with approximately 1 trillion total parameters and 49 billion active parameters, trained on 3,800 NVIDIA Grace Blackwell GPUs.
• The initiative aims to accelerate simulation cycles and enable closed-loop workflows connecting simulation, AI model development, and engineering design, with weights for Mistral Large 4 planned for release at the end of October.

Technical Integration

• Integration includes NVIDIA Omniverse libraries for GPU-accelerated 3D scene generation and SimReady asset creation.
• Benchmarks will assess agents' ability to use engineering tools, complete multistep workflows, and produce technically correct results while measuring execution time and computational cost.

Infrastructure and Background

• The collaboration builds on previous work including the co-development of Mistral NeMo and the NVIDIA Nemotron Coalition for frontier base models.
• Mistral is expanding NVIDIA-powered AI infrastructure to support a compute roadmap targeting 1 GW of capacity by 2030.