Microsoft and Mistral AI ink multi-billion dollar European infrastructure deal

Microsoft and Mistral AI announced a major expansion of their strategic partnership on July 21, 2026, centered on a multi-billion dollar investment in European AI infrastructure. The deal commits Microsoft to leveraging Mistral's expanded GPU capacity, which includes thousands of NVIDIA Vera Rubin GPUs — NVIDIA's next-generation accelerator designed for large-scale AI training and inference. This infrastructure investment aims to strengthen Europe's position in the global AI race while providing Mistral with the compute resources needed to scale its models.

As part of the partnership, Mistral's Medium 3.5 and OCR 4 models are now integrated into Microsoft Foundry and Copilot Studio. Medium 3.5 is a versatile language model optimized for enterprise tasks like summarization, reasoning, and code generation, while OCR 4 offers state-of-the-art optical character recognition for document processing. The integration allows enterprises to deploy these models across a range of environments — from cloud-based solutions to fully disconnected, on-premises setups — giving them greater control over data privacy and latency. This move underscores Microsoft's strategy to offer diverse AI model options and reinforces Mistral's role as a key European AI player.