Rate limiting

Each APEX key has a per-minute request budget, enforced with a fixed one-minute window. This is the APEX-specific limiter; the header conventions match the rest of the platform (see API conventions).

The budget

Headers

Every authenticated response carries the budget state:

X-RateLimit-Limit: 60
X-RateLimit-Remaining: 44
X-RateLimit-Reset: 1780735096

When you exceed it

Exceeding the budget returns 429 with a Retry-After header (seconds until the window resets) and the code RATE_LIMITED:

{
  "error": "rate limit exceeded",
  "code": "RATE_LIMITED",
  "message": "request budget for this API key is exhausted; retry after the window resets"
}

Rate-limited requests are still counted in per-key usage statistics, so operators can see who is hitting their budget. For high-volume workloads, prefer the batch estimate endpoint (up to 500 lanes in one call) over many single requests.