Performance & limits

Rate limiting

Rate limiting restricts how much work a caller or workspace can request within a period, helping keep demand within the service's supported capacity.

Also found under: HTTP 429, Throttling, Burst limit

How it works

A rate limit is different from a monthly quota and from the maximum size of one request. Bursts can hit a short-term limit even when monthly usage is low. Handle rejection explicitly, reduce unnecessary work and retry transient failures with bounded delay when appropriate.

In a Pubb integration

Pubb capacity can reject publications that exceed allowed rates. Inspect the response rather than treating every request as accepted, and preserve the idempotency key when retrying the same logical publication.

Explore plans and capacity

A practical example

A batch job spreads nonurgent notifications over time instead of publishing its entire backlog at once. Interactive updates retain capacity during the batch.