ISSUE:Selected model is at capacity. Please try a different model

Environment

  • Surface: CLI / IDE extension / App / Web / Cloud
  • Codex version: 26.905.11957
  • OS and IDE/browser, if relevant:
  • Model and plan or API-key authentication: PRO×20, every model

Bug

What happened, and what did you expect?
The PRO×20 account frequently encounters “Model Capacity” issues, making it nearly impossible to perform any development work. In contrast, the PRO×5 account works normally under the same conditions without similar capacity limitations. This issue is significantly impacting development and may indicate an account-level resource allocation, model capacity, or permission issue.

Reproduction

Steps or a minimal prompt/example:

Diagnostics

Exact error, relevant logs, or screenshot: Selected model is at capacity. Please try a different model.
When it happened (with timezone) and how often: very very frequently!!!

I have two Pro accounts. One keeps running into this issue. The other works fine, but I’ve used up 100% of its quota. This makes me wonder if the account has been flagged or banned?

No i dont think thats the case chatgpt has been having a lot of issues these lasst couple of weeks the platform has been just unstable snd broken theres issues every single day and i feel like they are getting worse

same ip and machine, i borrowed another account from my co-worker, logout my account and login borrowed account and it can work fine :cry:

It’s still the same on my end; things were fine for a while yesterday, but it went back to how it was afterwards.

Yes, this problem seemed to be resolved yesterday, but it has reappeared today.

sameSame problem here.!!!

Could this be related to the age of the account? Is it possible that newer Pro accounts are being flagged?

MAYBE! my pro5 is older than 20

I can only use the free codebuddy to work and codex cost me 200 dollars and can’t work.

Same here, just updated to 20x from plus and everything went from no problem to invisible wall.

It sounds like “you have done nothing wrong” but there is nothing we can do.

I am forced to use 5.5. Even 5.5 run into "at capacity once or twice per hour.

It isn’t a fix, but I am having an urgent issue that needs to be resolved. So, I’ve started a 60 second loop in another window to force goal to resume. I still need to babysit it for when it actually finishes the goal, but at least we are moving forward.

while sleep 60
tmux send-keys -t tmuxName:0.0 -l ‘/goal resume’
sleep 0.25
tmux send-keys -t tmuxName:0.0 Enter
end

I got the same problem!!!

My Codex keeps returning the error: “Selected model is at capacity. Please try a different model.” I am unable to get any work done.

Why is my codex always showing the message “Selected model is at capacity. Please try a different model.”? This is completely useless for doing any work. If this continues, I’m going to cancel my subscription to ChatGPT.

Same. Reproduces on 3 pro 20x accounts at same time. Makes gpt-5.6-{luna,terra,sol} and gpt-6.0-astra unusable at all. gpt-5.5 sometimes works (but really rarely).

Tried from 4 different IPs and 3 regions. While that, gpt-5.6-luna works fine on another free account.

9b6e2b3c-fafa-410a-8914-28b88cd3b2fa

Hey everyone, We are looking into this issue and will provide an update soon. Thanks for flagging this in community.

Same thing here. Pro 20x just unusable, since Astra launch. 5 days trying to develop paying $200 a month, I’m about to cancel this if does not get solved

This issue has severely impacted my work. Codex has been unusable for me, and I haven’t been able to get any work done today because of it. I really hope this can be resolved as soon as possible.

I’m using current account since gpt3, so I’m not really assuming it get flagged only due to timezone or proxy whatever. I am assuming however about cancelling pro service for this prolonged catastrophe without even a sign

recovered for like 2-3 rounds and fell demise hitherto