Skip to main content

Overview

caBRAPI applies request limits (rate limit) to guarantee stability, security and performance of the platform for all customers.
  • Limits are applied per contracted plan and measured in a fixed 1-minute window.
  • Each route has an independent counter — consuming one endpoint does not affect the limit of the others.
  • Usage is counted per API credential (your apiKey) — each integration measures its own usage.

Limits per plan

Basic

Up to 60 requests per minute.

Professional

Up to 120 requests per minute.

Hyper

Up to 1,600 requests per minute.
If the limit is exceeded, new requests are temporarily blocked, returning a 429 error until the window restarts.

Control headers

Every response that passes authentication and plan validation includes headers that allow tracking consumption in real time (early 401, 403, or 404 responses abort before rate-limit headers are applied): When the limit is reached, the error response also includes:

When the limit is exceeded

When exceeding the limit, the API responds with 429 Too Many Requests and refuses new requests until the window restarts:
The response uses the same code field as the other API errors. See Errors.
The wait time is informed in the Retry-After header (in seconds), not in the response body.

Best practices

  • Monitor the X-RateLimit-Remaining header to self-regulate before reaching the limit.
  • When receiving a 429, respect the Retry-After and only try again after the indicated time.
  • Implement retry with exponential backoff to handle traffic peaks.
  • Avoid aggressive polling: query data at larger intervals or use webhooks to receive events in real time.
If your flow demands more requests, consider a plan upgrade — the upgrade raises the limit immediately.

Pagination

In listing endpoints, use the page and limit parameters to control the volume of data per response.
  • page: integer greater than or equal to 1
  • limit: integer between 1 and 100
When these parameters are invalid, the API returns:
  • 400 INVALID_PAGINATION — when page and cursor are combined
  • 400 INVALID_PAGINATION_LIMIT
  • 400 INVALID_PAGINATION_PAGE
  • 400 INVALID_PAGINATION_CURSOR — when cursor is invalid or malformed

Cursor pagination

In large lists (more than ~100 items), prefer cursor pagination, which keeps performance at any data volume:
  • Make the first call without cursor and use the nextCursor field from the response.
  • Send the nextCursor as cursor in the next request to advance in the list.
  • The end of the list is indicated by nextCursor: null + hasMore: false (the API responds 200 with an empty array).
  • In offset mode, responses also include nextCursor/hasMore, allowing migration to cursor without starting from scratch.
Limits may be adjusted according to new plans or specific conditions of caBRAPI. Always check the updated documentation to ensure correct information.