Rate limits
The limits, the headers that report them, and how to stay under them.
Limits are per workspace, not per key, so adding keys does not add capacity.
Headers
Every response carries the current state:
When you are limited
A 429 includes Retry-After in seconds. Wait that long. Retrying sooner extends the window rather than shortening it.
Staying under
Read in batches with pageSize rather than one record at a time, cache what does not change often, and spread scheduled jobs instead of starting them all on the hour.
Higher limits
Limits scale with your plan. If a workload legitimately needs more than the plan allows, that is a conversation rather than a workaround.
Actualizado el 28 ago 2026