Regional inference now available on AI Gateway
By Steven Van ·
AI Gateway's new inferenceRegion setting pins requests to a US or EU data center, with a failure instead of a silent fallback if no provider can serve it.
Vercel's AI Gateway now supports regional inference. Setting inferenceRegion on a request pins it to a US or EU data center, and any data the provider stores at rest is kept in that same region. Previously, teams with data residency requirements had to configure regional routing separately for each provider, with no reliable way to confirm where a request actually ran.
- Requests are pinned via providerOptions.gateway.inferenceRegion, with a zone scope of us or eu, or left as global to route to whichever region is available.
- If no provider can serve the chosen region, the request fails rather than running elsewhere.
- Every response reports the region that served it, so the resolved location can be confirmed after the fact.
- The model list can be filtered for US or EU availability, or checked via the regions array on /v1/models.
Regional routing is opt-in: without inferenceRegion set, requests route globally with no residency guarantee. Pinning a region can also raise the price, since the provider sets the regional rate, often around 10% above standard, and Vercel passes it through without adding a markup. Read the full regional inference documentation for per-provider overrides and BYOK behavior.