Mistral released regional inference capabilities and open model access, saying it enables enterprises to control where AI runs and retain data residency. It is the company's first major update to its sovereign AI infrastructure since launching its flagship models in 2024.
Mistral reported 1 GW of planned compute capacity by 2030, measured across its European infrastructure. That compares with 500 MW of capacity announced in 2025.
Regional inference is built on Mistral's cloud-native platform and targets compliance with data residency and regulatory requirements. Availability begins with general release, initially for European enterprises and institutions.
"Mistral Regional Endpoints let customers choose whether their inference runs in Europe or the US, helping them align inference location with their data-residency, regulatory, and latency requirements," said Matan Grinberg, CEO and cofounder of Factory. Mistral is the only European AI lab to offer both: choice of processing region and a committed, SLA-backed service level.
The announcement follows Mistral's expansion into open model contributions. The company emphasized the importance of sovereign infrastructure for Europe's technological independence.
Mistral did not say how it will finance the 1 GW capacity build, and raised the open question of how to ensure long-term compute availability. The company said it will continue to expand its coalition of enterprises and institutions to secure compute capacity.
Source: mistral
