{% extends "_layout.html" %} {% set title = "Too many requests" %} {% set status = "429 Too Many Requests" %} {% set description = "Calls per unit time. Honour Retry-After." %} {% block main %}

You are calling this deployment faster than it accepts. This is about how often you ask, not about what your account may hold or consume.

What to do

  1. Wait for Retry-After, then send the request again.
  2. If you poll, poll at the interval the server gives you: a job read answers with its own Retry-After.

The body

{
  "type":     "https://sc.example.com/server-errors/rate-limited",
  "title":    "Too many requests",
  "status":   429,
  "detail":   "...",
  "instance": "/v1/jobs/01J9...",
  "trace_id": "4bf92f3577b34da6a3ce929d0e0e4736"
}

detail is prose and may be reworded at any time. Branch on type — never parse detail.

Where it comes from

any endpoint The deployment’s own request-rate ceiling.
{% include "_help.html" %}

At the two OAuth endpoints

Raised at POST /v1/auth/token and POST /v1/auth/device too, before any OAuth processing, and in problem+json there as everywhere — never as an OAuth error. A client handles it by its status and type, and waits Retry-After.

What this is not

limit-exceeded is a published ceiling on what you hold, and names it in limit. This one has no key to name.

{% endblock %}