Source: OpenAI API data residency guide

OpenAI's regional API domains: ten storage regions, of which only some run inference in-region; per-project or per-request selection and a 10% uplift for newer models.

Contents
  • What: official API documentation
  • Publisher: OpenAI
  • Date: accessed 2026-11-14
  • Original: Data residency guide

Summary

Distinguishes regional storage (customer content at rest) from regional processing (inference). Many regions are storage-only, and several features fall back to out-of-region processing.

Key claims

  • Storage and processing in-region: US (us.api.openai.com), Europe/EEA+Switzerland (eu.), UAE (ae., limited models, contact sales); Australia, Japan, India process in-region only for gpt-6-luna (Chat Completions and Responses, Standard pay-as-you-go, no hosted tools; India also via sales).
  • Storage-only (inference may run elsewhere): Canada, Singapore, South Korea, UK.
  • Region is fixed per project at creation; a Global project created with “per-request data residency” can instead target a regional domain per request.
  • Models released on or after 2026-03-05 carry a 10% price uplift on data-residency endpoints.
  • Caveats: store=true on chat completions is not allowed outside the US; in regions without regional processing, extended prompt caching may process and temporarily store data outside the region; files and vector stores are storage-only; Realtime tracing is not EU-residency compliant; system data is excluded.

Feeds

PII and data privacy

Suggest changes →