Rate Limits

Rate limits are enforced per workspace, so adding more API keys does not raise your ceiling. Sustained limits depend on your plan:

A per-key burst limit of roughly 3 requests / second also applies on top of the sustained per-workspace limit.

Rate limit headers

Every response includes rate limit headers so you can pace your requests:

X-RateLimit-Limit: 600
X-RateLimit-Remaining: 597
X-RateLimit-Reset: 42          # seconds until the window resets

Exceeding the limit

If you exceed a limit, you'll receive an HTTP 429 Too Many Requests. A 429 additionally includes a Retry-After header telling you how many seconds to wait:

Retry-After: 42

The body uses the standard error envelope:

{
  "error": {
    "type": "rate_limit_error",
    "code": "throttled",
    "message": "Request was throttled. Expected available in 42 seconds."
  },
  "request_id": "9f2c1a7e5b8d4f31a0c6e2d9b7a4f108"
}

It's important to handle a 429 response. We recommend implementing retries with an exponential backoff strategy, and respecting the Retry-After header when present.

If your use case requires a higher limit, contact us at sales@audienceful.com and we'll be happy to discuss your needs.