OutageRadar

Checking… Checking…

OpenAI API error 429: rate limit reached or slow_down

What this OpenAI API error means, whether it is on OpenAI’s side or yours, and what to do now. All times are UTC.

Is it them or you?

Request limit: your account’s pace

This error is about your account’s request limits. OpenAI’s servers answered and turned the request away, which usually means they are up. The live status below helps rule out a wider problem.

Checking…

ChatGPT / OpenAI

Official status: checking…

Our check: checking…

What it means

OpenAI returns a 429 when you send too many requests or tokens in a short time (“Rate limit reached for requests”).

It also returns a 429 with type rate_limit_error and code slow_down when your request rate rises faster than the service can safely handle. OpenAI notes this can happen even when you are within your per-minute limits.

Billing and quota problems use status 429 too, but they are different: check error.code. See the OpenAI quota page.

What to do

  1. If a Retry-After header is present, wait at least that long. If it is missing, use exponential backoff with jitter and limit the number of retries.
  2. For slow_down: reduce your request rate, then increase it gradually, and keep your traffic steady.
  3. Limits apply per organization, not per user, so check what the rest of your team is sending.
  4. OpenAI’s Python library retries some errors itself. If you also retry in your own code, respect the original delay and account for the library’s retries.

Retry it safely in your own code

You may not need this: the official SDKs retry on their own (see above). If you call the API directly, this helper waits and retries, using the server’s requested delay when you pass one. Treat rate_limit_error (including slow_down) as retryable; never retry billing or quota codes.

Python

import random
import time


def with_retries(call, is_retryable, retry_after=lambda err: None, attempts=5):
    """Run call(). On a retryable error, wait and try again, up to attempts tries.
    retry_after(err) returns the server's requested wait in seconds, or None."""
    for attempt in range(attempts):
        try:
            return call()
        except Exception as err:
            if attempt == attempts - 1 or not is_retryable(err):
                raise
            wait = retry_after(err)
            if wait is None:
                wait = min(60, 2 ** attempt) + random.random()  # backoff + jitter
            time.sleep(wait)

JavaScript

async function withRetries(call, isRetryable, retryAfter = () => null, attempts = 5) {
  // retryAfter(err) returns the server's requested wait in seconds, or null.
  for (let attempt = 0; ; attempt++) {
    try {
      return await call();
    } catch (err) {
      if (attempt >= attempts - 1 || !isRetryable(err)) throw err;
      let wait = retryAfter(err);
      if (wait == null) wait = Math.min(60, 2 ** attempt) + Math.random(); // backoff + jitter
      await new Promise((resolve) => setTimeout(resolve, wait * 1000));
    }
  }
}

Recent ChatGPT incidents

History logged by OutageRadar.

Full history: Is ChatGPT down?

Source: Error codes (OpenAI docs) · Rate limits (OpenAI docs). Last checked against official docs: Nov 10, 2026. If the docs change, they win.