Rate limits

Each account gets 500 requests a day. The allowance belongs to the account, not to any one key, and the API counts it for you.

The allowance

  • 500 requests per account per day.
  • Shared by every key the account owns. A second key does not buy a second allowance.
  • It resets at midnight UTC, not 24 hours after your first request.
  • Every request that carries a key counts, including ones that answer with an error.

Your dashboard shows what each key has used, per day and per endpoint.

Going over

Once a key is out of requests, every call with it answers 429 until the window moves:

{
  "error": {
    "message": "Rate limit exceeded for this API key. Try again later."
  }
}

There is no Retry-After header yet. Everything is available again at midnight UTC; until then, back off rather than retrying immediately.

Staying under

  • Cache what does not change. /vocabularies changes rarely; fetch it once at start-up, not per request.
  • Ask for less. limit and the fields parameter on search both cut the work.
  • Use a key per application anyway. It will not raise the ceiling, but the dashboard then tells you which one is spending it.
  • Need more than 500 a day? Get in touch and we will raise it.