Rate limits

The API allows 600 requests per minute per workspace. All API keys of a workspace share that budget; separate workspaces have separate budgets.

When you exceed it, the API responds 429 Too Many Requests with a Retry-After header (seconds) and a message that says how long to wait:

{  "status": "error",  "message": "Too many API requests in a row. Please retry after 12 seconds"}

Handling 429

• Wait the number of seconds in Retry-After, then retry. The message carries the same number.
• Spread bulk jobs (retraining many sources, exporting conversations) over time instead of firing them at once.
• Some endpoints also have their own, stricter limits; they answer with the same 429 shape.

Try-it requests from these docs

Requests you send from these pages are normal API calls and count toward the workspace budget. Signed-in requests are additionally limited to 60 per minute per user, so the docs can never use up the whole workspace budget.