In-region inference, open models, and new European infrastructure for sovereign AI.
mistral ·
Mistral AI has unveiled a comprehensive strategy to strengthen AI sovereignty by giving enterprises and governments greater control over their artificial intelligence infrastructure and data. The company is now offering generally available regional inference endpoints that allow clients to route processing either within Europe or the United States to satisfy local regulatory and latency requirements. Alongside this, a newly introduced priority tier provides guaranteed uptime and customized rate limits for mission-critical applications. Expanding beyond its proprietary offerings, Mistral will also host third-party open-source models, beginning with Z.ai’s GLM-5.2, ensuring they operate under identical regional safeguards and service guarantees. To tackle the growing shortage of reliable computing resources, Mistral is uniting major European corporations to aggregate multi-year demand into European Compute Units. These pooled commitments will directly finance the development of up to one gigawatt of dedicated AI infrastructure across Europe by 2030. Ultimately, the initiative aims to establish a resilient, locally governed ecosystem where organizations can deploy advanced AI without compromising data residency or strategic autonomy.
mistral