Mistral made regional endpoints generally available, letting customers choose whether their inference runs in Europe or the US to match data-residency and latency requirements. It also opened a public preview of a priority tier with custom rate limits and an uptime guarantee, and began hosting outside open models starting with Z.ai's GLM-5.2. Some processing by sub-processors may still occur outside the chosen region.
What changed
Customers could not choose the region their inference ran in on Mistral's own platform, had no SLA-backed service tier, and could only run Mistral's own models there.
What it unlocks
Running inference on Mistral's platform with processing kept in either Europe or the US, with a committed uptime guarantee, and running a third-party open model (Z.ai's GLM-5.2) under the same controls.
- up to 1 GW of capacity by 2030
What you need to act on it
- Mistral platform account
- priority tier is in public preview
- mistral.ai2026-08-11