Hermeseus Docs
Legacy docs

Rate limits

Requests are rate limited per access token. Every response tells you how much of your budget remains, and a request over the limit returns 429.

Headers

Each response carries your current budget:

HeaderMeaning
X-RateLimit-LimitRequests allowed per minute.
X-RateLimit-RemainingRequests left in the current window.
Retry-AfterSeconds to wait, present on a 429.

When you hit the limit

A throttled request returns 429 Too Many Requests with the standard error body:

{
  "error": {
    "type": "rate_limit_error",
    "code": "rate_limited",
    "message": "Too many requests. Slow down and retry.",
    "request_id": "req_…"
  }
}
Back off and retry. Wait for the Retry-After interval, then retry. Build exponential backoff into your client so a burst smooths out instead of hammering the limit.