Error responses
Errors include a readable explanation and a code:errorCode to handle errors in your integration. reason explains the problem in plain language.
Request and cache headers
Responses includeX-Request-ID for troubleshooting and Cache-Control: private, no-store. You may supply a UUID in the X-Request-ID request header.
Include the response request ID and endpoint when contacting support. Never include your key or authorization header.
Request limits
The standard limits are 60 requests per minute and 2,000 per day, shared by your integrations. Check the response headers for your current allowance. Limits reset at the start of each UTC minute and at midnight UTC. Replacing your key keeps the same allowance. Requests with a valid key and active Plus count toward both limits, including requests that return errors or429. Requests rejected for authentication or Plus access do not count. Wait before retrying to avoid using up your allowance.
Read the quota headers
Quota headers describe your remaining allowance:
For example,
Retry-After: 42 means wait at least 42 seconds before another attempt. If both limits are reached, wait for the later reset.
Quota headers may be absent on authentication errors or temporary service failures.
Creating or replacing a key in the app has a separate limit of five attempts per minute.