Mistral AI has announced a three-part push on European AI sovereignty, covering regional inference control, third-party open model access, and a coalition to secure long-term compute capacity, the company said on 11 August 2026.
Mistral Regional Endpoints, now generally available, let customers choose whether inference runs in Europe or the US, while a new Priority Tier in public preview offers committed service levels and an uptime SLA for critical workloads. Mistral said it is the only European AI lab offering both regional choice and an SLA-backed service tier.
The company is also opening its platform to third-party open models, starting with Z.ai's GLM-5.2, which will run under the same regional controls and service commitments as Mistral's own models. Factory chief executive Matan Griberg said the approach let his company maintain data residency and compliance requirements while supporting open source.
Separately, Mistral is forming a coalition of enterprises, including ASML, CMA CGM, Amadeus and Caisse des Dépôts, whose multi-year commitments will fund European AI infrastructure through a new mechanism called European Compute Units. Mistral is targeting up to 1 gigawatt of capacity by 2030. Executives from the participating companies framed the initiative as central to European technological independence and long-term competitiveness.
