Rate limits

Ceilings

Headers

Every response carries X-RateLimit-Limit and X-RateLimit-Remaining. A 429 also carries Retry-After, in seconds. Read those headers rather than counting your calls: the ceiling can change with the plan, at runtime.

HTTP/1.1 429 Too Many Requests
X-RateLimit-Limit: 60
X-RateLimit-Remaining: 0
Retry-After: 37

{ "error": { "code": "rate_limited",
    "message": "Rate limit of 60 requests per minute exceeded." } }

Per key, not per account

Two integrations on the same account therefore do not starve each other, and you can tell which one is consuming. It is also what makes revoking a runaway key immediately useful to the rest of the account.

On a 429

Wait the number of seconds given by Retry-After, then retry. The window is fixed: past that delay the counter is back to zero.