[Windows][Codex App 26.826.12353] GPT-5.6 Sol repeatedly returns “Selected model is at capacity”
Nobody has claimed this yet.
- Dominant language
- Rust
- Stars
- 125k
- Forks
- 19.4k
- PR merge metrics
- PR metrics pending
Description
What version of the Codex App are you using (From “About Codex” dialog)?
26.826.12353
What subscription do you have?
ChatGPT Plus
What platform is your computer?
Windows x64
What issue are you seeing?
GPT-5.6 Sol in Codex Desktop repeatedly fails with the following error:
Selected model is at capacity. Please try a different model.
The error occurs repeatedly and prevents normal use of GPT-5.6 Sol in Codex Desktop. Retrying does not reliably recover the request.
This appears to be a model-serving capacity / routing problem rather than a local project-specific error, because the failure is surfaced directly as a selected-model capacity error.
What steps can reproduce the bug?
-
Open Codex Desktop on Windows.
-
Use Codex App version
26.826.12353. -
Select
GPT-5.6 Sol. -
Send a normal prompt.
-
Codex returns:
Selected model is at capacity. Please try a different model. -
Retry the request.
-
The same capacity error occurs repeatedly, preventing normal use of the model.
Feedback ID: 01a02ea7-69dd-7cb0-9a46-273277a91e63
What is the expected behavior?
GPT-5.6 Sol should accept and process normal Codex requests when the account still has access to the model.
If the backend is temporarily at capacity, Codex should ideally retry automatically with bounded backoff and preserve the current task/session state instead of repeatedly terminating the request and requiring manual retries or a model switch.
Additional information
- Codex Desktop version:
26.826.12353 - Subscription: ChatGPT Plus
- Platform: Windows x64
- Selected model: GPT-5.6 Sol
- Feedback ID:
01a02ea7-69dd-7cb0-9a46-273277a91e63 - Observed on: September 2, 2026 (UTC+8)
Similar recent reports appear to include #41790 and #41805, but this report provides a Windows Codex Desktop reproduction on version 26.826.12353 together with a fresh Feedback ID for diagnostics.
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Begin with the Windows x64 reproduction using Codex Desktop 26.826.12353, GPT-5.6 Sol, and the supplied Feedback ID. Determine whether the repeated capacity response is produced by local routing or the model-serving backend, then verify that a bounded retry preserves the task/session state and handles temporary capacity errors without requiring repeated manual retries.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- rust
- Domain
- api, backend
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Active
- Clarity
- Mostly clear
- Newbie friendliness
- 35/100