Selected model is at capacity
What version of the Codex App are you using (From “About Codex” dialog)?
26.422.30944
What subscription do you have?
Pro 200 USD
What platform is your computer?
Microsoft Windows NT 10.0.26200.0 x64
What issue are you seeing?
Message sayung Selected model is at capacity, even though my token limits are only used 10% of my plan.
I’ve had a session work to test an electorn app’s interfaces and log each window’s errors, and suggest optimization improvements based on log outputs and delays, it Worked for 3h 9m 39s, then stopped suddenly saying Selected model is at capacity.
Then every conversation with GPT 5.5 xhigh on a Pro 200 USD account kept saying the same thing, so i was no longer able to use this model so i had to switch back to 5.4 xhigh to continue my work.
Is it an issue from my side or is this an OpenAI issue? I’ts been2 hours and still the same issue.
What steps can reproduce the bug?
Running 5.5 xhigh on a long running simulation, testing and documentation session (3 hours +).
What is the expected behavior?
Continue working until limit exhaustion or finalization
Additional information
_No response_
44 Comments
Potential duplicates detected. Please review them and close your issue if it is a duplicate.
Powered by Codex Action
The issue you mentioned is in CLI, but my issue is with codex, also i'm on the highest tier plan.
This occured for a while, then it stopped.
me too
Me too
If you receive this message, it does not mean that you are being rate limited or are out of usage. It means that we are running low on capacity to serve this model in your region.
me too
me too
Happening for me on 5/26:
same on 6/9
same at 6/9
I've encountered this problem too, and I'd like to ask how to solve it effectively. Will it just get better after waiting for a while?
This issue isn't resolved. Codex CLI should retry with a backoff instead of aborting what it was working on...
The Codex app sadly also doesn't retry, but I guess this is a measure to shed load. Still quite annoying.
Pro 100 USD
ubuntu22.04 vsode+codex
GPT-5.5 High
Selected model is at capacity. Please try a different model.
I'm also experienced this today...
Same here. Not fun at all!
Happen on me today. In both remote session and local session. BTW, it might work for just 2mins but stop sliently again.
Pro 200 USD
Linux & Macos
GPT-5.5 high
Selected model is at capacity. Please try a different model.
Damn!!! GPT-5.5-xhigh also be declined too.
me too
Pro 5x
Linux Debian
GPT-5.5 High
Selected model is at capacity. Please try a different model.
Switched to Medium it seems going. But I'm a little bit disappointed, I've just subscribed the Pro plan guys
Edit after 13 minutes: gpt-5.5 medium is stuck working:
Falling back from WebSockets to HTTPS transport. stream disconnected before completion: An error
occurred while processing your request. You can retry your request, or contact us through our help
center at help.openai.com if the error persists. Please include the request ID 922b2bc0-65f7-45d4-
b24f-b1a1ba30ab54 in your message.
I ran into this problem today as well. has anyone managed to solve it?
5.5-xhigh 、codex app、pro-20x
Also me, 5.5 x-high fast, codex windows app, pro-20x
Same here, 5.5 xhigh/high, codex mac app, pro-20x
same issue multiple times per day , please fix it.
same, they might have to much success with anthropic drama
me too
me too
Same issue
me too
same for me
me too
I got this too, please fix it!
Same issue
Please upvote the issue instead of adding more comments; otherwise, it spams everyone else.
+1 codex
me too
me too
+1. Switched to GPT 5.4, working fine now.
me too in 5.5, but I switched to GPT 5.4 and worked fine with any problem
eu também
me too in 5.5;
5.4 does not show the problem and it is much faster at the moment. not felling any diff between both.
me too
This issue has been mitigated.
I'm going to close and lock this issue because it's quite old. If you see this problem again, please open a new issue.