GPT-5.5 capacity error interrupts ongoing image-based sessions on Windows app 26.623.61825

Open 💬 7 comments Opened Jun 29, 2026 by gizemo8318-design
💡 Likely answer: A maintainer (github-actions[bot], contributor) responded on this thread — see the highlighted reply below.

What version of the Codex App are you using (From “About Codex” dialog)?

Version 26.623.61825 • Released 2026. 6. 29.

What subscription do you have?

ChatGPT Plus

What platform is your computer?

x64

What issue are you seeing?

019f136d-3165-79e2-aad8-e275a1a81023 Similar to #19583, but recurring on version 26.623.61825.

  1. Open the Codex desktop app on Windows.
  2. Select GPT-5.5.
  3. Start or continue a conversation involving image analysis.
  4. Continue working in the same thread for several turns.
  5. After initially working normally, the thread suddenly fails with:

"Selected model is at capacity. Please try a different model."

  1. Retrying in the same ongoing thread may continue to fail, interrupting the work and its conversational context.

No code snippet is required because this is a Codex desktop app/session issue.

Feedback ID: 019f136d-3165-79e2-aad8-e275a1a81023
Similar to #19583, but recurring on app version 26.623.61825.

What steps can reproduce the bug?

Feedback ID: 019f136d-3165-79e2-aad8-e275a1a81023

What is the expected behavior?

_No response_

Additional information

_No response_

View original on GitHub ↗

7 Comments

github-actions[bot] contributor · 21 days ago

Potential duplicates detected. Please review them and close your issue if it is a duplicate.

  • #30569
  • #30575
  • #30547
  • #30546
  • #30073

Powered by Codex Action

gizemo8318-design · 21 days ago

This appears to be the same capacity incident, but my report includes a diagnostic Feedback ID from the Windows Codex app. I will leave this issue open for maintainers to review and consolidate if appropriate.

QianMuXiao · 21 days ago

My shorter conversation threads seem to be fine; the issue only appeared in one particularly long thread that had undergone multiple rounds of compression. Could it be that OpenAI is running low on hardware resources for long-context processing?

bjicn · 21 days ago
我较短的对话线程似乎很好;这个问题只出现在一个经过多轮压缩的特别长线程中。可能是OpenAI用于长上下文处理的硬件资源不足吗?

No, this problem still occurs after I open a new thread, and sometimes it will be interrupted and this problem will occur during the answer process.

QianMuXiao · 21 days ago
> 我较短的对话线程似乎很好;这个问题只出现在一个经过多轮压缩的特别长线程中。可能是OpenAI用于长上下文处理的硬件资源不足吗? No, this problem still occurs after I open a new thread, and sometimes it will be interrupted and this problem will occur during the answer process.

Well, I’ve noticed that my short conversations still seem to work occasionally, but that primary, long-running conversation has become completely unusable. I’ve started a new conversation thread in the project directory and set it up to directly read all the previously saved dialogue.

LagodlV · 21 days ago

Same problem here. Just working in small conversations but not all the time.

> > 我较短的对话线程似乎很好;这个问题只出现在一个经过多轮压缩的特别长线程中。可能是OpenAI用于长上下文处理的硬件资源不足吗? > > > No, this problem still occurs after I open a new thread, and sometimes it will be interrupted and this problem will occur during the answer process. Well, I’ve noticed that my short conversations still seem to work occasionally, but that primary, long-running conversation has become completely unusable. I’ve started a new conversation thread in the project directory and set it up to directly read all the previously saved dialogue.
QianMuXiao · 20 days ago
Same problem here. Just working in small conversations but not all the time. > > > 我较短的对话线程似乎很好;这个问题只出现在一个经过多轮压缩的特别长线程中。可能是OpenAI用于长上下文处理的硬件资源不足吗? > > > > > > No, this problem still occurs after I open a new thread, and sometimes it will be interrupted and this problem will occur during the answer process. > > > Well, I’ve noticed that my short conversations still seem to work occasionally, but that primary, long-running conversation has become completely unusable. I’ve started a new conversation thread in the project directory and set it up to directly read all the previously saved dialogue.

It was restored last night, so I can continue with my main conversation—plus, I received a welcome reset of my free quota.