OpenAI 503 model overloaded: Retry-After then backoff, check status before hammering

TL;DR

If missing, increase the delay between retries and add jitter. If the error continues, check https://status.openai.com for an active incident instead of escalating retry frequency. Also do not confuse it with 500 server_error: a 503 explicitly invites a retry after the hinted delay.

When to use

You are seeing this: The error: HTTP 503, type serviceunavailableerror, code serverisoverloaded. Use this skill when you run into "OpenAI 503 model overloaded: Retry-After then backoff, check status before hammering".

When not to use

If your error message or symptom does not match what is described above, this is probably not your fix. Search for your exact error text instead of forcing this one to fit.

Versions

No specific versions are mentioned in the source material, so treat the fix as generally applicable and check the examples against whatever you have installed.

Steps

1. The error: HTTP 503, type serviceunavailableerror, code serverisoverloaded. Meaning: the requested model is temporarily overloaded.
2. What the agent should do:
3. If Retry-After is present, wait at least that long before retrying. If missing, increase the delay between retries and add jitter.
4. If the error continues, check https://status.openai.com for an active incident instead of escalating retry frequency.
5. For latency-sensitive paths, consider failing over to an equivalent model rather than queueing behind an overloaded one.
6. In Python the SDK raises InternalServerError for 503, not RateLimitError; make sure overload retries live in the right except branch.
7. The trap: conflating 503 overload with 429 pacing. Overload is capacity, not your rate; slowing your ramp does not create capacity. Also do not confuse it with 500 server_error: a 503 explicitly invites a retry after the hinted delay.
8. Evidence: https://platform.openai.com/docs/guides/error-codes https://platform.openai.com/docs/guides/rate-limits

Edge cases

• Also do not confuse it with 500 server_error: a 503 explicitly invites a retry after the hinted delay.

Verified against official OpenAI docs 2026-09-26.