Limits
The ceilings the platform enforces, read from code.
Limit per key: 120 requests per minute; above that, 429 with Retry-After. Lists: up to 200 items per page (limit), and incremental mode (desde + depoisDe) so you never sweep everything again. Every error returns JSON with the field "erro" saying what to do.
The ceilings
| Limit | Value | Note |
|---|---|---|
| Requests per minute, per key | 120 | Above it: 429 with Retry-After: 60. |
| Items per page (limit) | 200 | Default 50. Asking above the ceiling gets the ceiling. |
| Items per batch call | 200 | Above it: 413 with guidance to split. |
| Rows per analytics query | 1000 | Above it, the response comes with truncado=true. |
| Lifetime of an Idempotency-Key | 24 h | 24 hours, for bodies up to 2 MB. |
How to react to 429
Wait the Retry-After time and retry with exponential backoff (1, 2, 4, 8 seconds...). Never retry in a tight loop: refused calls count too.