Is anyone finding that GPT-5 is timing out on longer tasks? I’ve burned a decent amount of tokens having it work through some longer feature implementation on a codebase, and after a while it just errors out with a timeout.
5 Likes
For longer tasks, try the background mode.
You will receive a response ID that you can poll for answer or be notified with a webhook, without the risk of losing it.
https://platform.openai.com/docs/guides/background#polling-background-responses
1 Like
Thanks – I’m using within Xcode and so, a bit limited in terms of how it’s deciding to connect under-the-hood.
2 Likes
same issue here. Very unfortunate.
Same issue using 5-mini
My application prompt takes about 5 minutes to complete.
Someone on my stream is timing out the socket.
Not acceptable for an application ???
Absolutely. Its great when you can get it to work. But it takes several back and forths to get it to obey and to not crash. Massive amount of time wasted on ChatGPT Pro!
Even with background mode, it sometimes takes me 3+ hours to get a response from gpt-5-nano. or gpt-5-mini I’ve been having this issue since last week when I started using the API.
2025-11-04 10:26:52,427 - INFO - HTTP Request: GET https://api.openai.com/v1/responses/resp_0f677aa8add8416200690a17f7abb08191843a20393423c464 "HTTP/1.1 200 OK"
2025-11-04 10:26:52,427 - INFO - [BG] Current status: in_progress | elapsed=11637.2s ... polling again in 10s
2025-11-04 10:27:03,935 - INFO - HTTP Request: GET https://api.openai.com/v1/responses/resp_0f677aa8add8416200690a17f7abb08191843a20393423c464 "HTTP/1.1 200 OK"
2025-11-04 10:27:03,935 - INFO - [BG] Current status: in_progress | elapsed=11648.8s ... polling again in 10s
2025-11-04 10:27:14,175 - INFO - HTTP Request: GET https://api.openai.com/v1/responses/resp_0f677aa8add8416200690a17f7abb08191843a20393423c464 "HTTP/1.1 200 OK"
2025-11-04 10:27:14,176 - INFO - [BG] Current status: in_progress | elapsed=11659.0s ... polling again in 10s
1 Like