Regional AI inference allows Slack customers to have large language model (LLM) calls for native Slack AI features routed to a supported region of their choosing. Starting in November 2026, customers in Japan can contact Slack support to request regional AI inference.
What to expect
All model calls (prompts sent to LLMs and responses) are routed through your selected region. Data residency for Slack is separate from regional AI inference.
We don’t support cross-region fallback, so requests will fail if they can’t be processed in your designated region.
Regional AI inference is applied at the workspace or organisation level, and requests are always routed to your region regardless of individual members’ locations.
Available regions
Processing region
Regional endpoints
Japan
Osaka, Tokyo
Regional inference and Slack Connect
If you’re working with people from another organisation in a Slack Connect channel or direct message (DM), AI requests made by your members will be routed to your specified inference region. AI requests made by members of other organisations will be routed based on their workspace or Enterprise organisation’s configuration.
Who can use this feature?
Only the workspace primary owner (Business+) or the org primary owner (Enterprise+) can enable regional AI inference
