Issue Station
Sign In
Tools hereToolbox

codex - 💡(How to fix) Fix Codex App (Azure) intermittent stream disconnect + false content_filter; CLI works with same config [2 comments, 3 participants]

Official PRs (…)
ON THIS PAGE

Recommended Tools

×6

Utilities matched from this issue’s tags and category — try them while you read without losing context.

GitHub issue graph ai analysis

Paste a GitHub issue URL. We fetch that issue, discover linked issues from bodies/comments/timeline, collect linked pull requests, and produce a structured English report.

The report is written in English Markdown for sharing and archival.

Helpful · Quick feedback

Loading…
GitHub stats
openai/codex#17797•Fetched 2026-04-15 06:28:04
View on GitHub
Comments
2
Participants
3
Timeline
9
Reactions
0
Author
Sir-IanSir-Ian
Participants
etraut-openaietraut-openaigithub-actions[bot]github-actions[bot]Sir-IanSir-Ian
Timeline (top)
labeled ×4commented ×2closed ×1cross-referenced ×1
RAW_BUFFERClick to expand / collapse

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

26.409.20454

What subscription do you have?

Azure OpenAI (custom deployment via Foundry)

What platform is your computer?

Darwin 25.4.0 arm64 arm

What issue are you seeing?

This issue started after updating the Codex app last week. Prior to the update, the same workflows, prompts, and Azure configuration worked reliably without reconnect loops or content_filter errors.

Using Codex desktop app with Azure provider results in intermittent failures with:

  • "Reconnecting..."
  • "stream disconnected before completion"
  • "Incomplete response returned, reason: content_filter"
  • repeated "I'm sorry, but I cannot assist with that request."

This occurs even for harmless prompts (e.g., "hi") and normal workflows (e.g., generating slides from provided content + PDF). The response often starts correctly, then disconnects mid-stream and enters a retry/refusal loop. This behavior is intermittent but frequent and reproducible across multiple models and over multiple days.

What steps can reproduce the bug?

Uploaded thread: 019d8ca1-ca17-7732-9de9-5d659f9a4f56

What is the expected behavior?

The Codex desktop app should behave consistently with the CLI when using the same Azure provider, config, and model. Harmless prompts and standard workflows should not trigger content_filter errors or stream disconnects. Responses should complete without reconnect loops or fallback refusal messages.

Additional information

This behavior did not occur prior to updating the Codex app last week. The same setup (Azure provider, same models, same prompts, same workflows) worked reliably before the update.

No changes were made to:

  • Azure deployments
  • API keys or permissions
  • config.toml (outside of debugging attempts after issue began)

This suggests a regression in the Codex app related to:

  • streaming / reconnect handling
  • model selection / config state
  • or interaction with Azure content filtering

extent analysis

TL;DR

The issue is likely related to a regression in the Codex app's streaming or reconnect handling, and a potential workaround could be to try rolling back to a previous version of the Codex app or adjusting the Azure content filtering configuration.

Guidance

  • Review the changelog for version 26.409.20454 to identify potential changes that could be causing the regression.
  • Verify that the Azure content filtering configuration is correctly set up and not blocking the responses from the Codex app.
  • Test the same workflows and prompts using the CLI to confirm if the issue is specific to the desktop app.
  • Check the Codex app's configuration files (e.g., config.toml) for any changes that could be contributing to the issue, even if no intentional changes were made.

Example

No code snippet is provided as the issue does not explicitly mention a specific code-related problem.

Notes

The issue seems to be related to a regression in the Codex app, and the exact cause is still uncertain. Further investigation is needed to determine the root cause of the problem.

Recommendation

Apply workaround: Given the intermittent nature of the issue and the fact that it started after an update, it might be worth trying to adjust the Azure content filtering configuration or rolling back to a previous version of the Codex app to see if the issue persists. This could help mitigate the problem until a more permanent fix is available.

Vote matrix · Quick signals

Works
Did the solution work? Tap to confirm.
Easy Fix
Was it a quick fix?
Time Saver
Did it save you time?
Blocking
Was it severely blocking?
Common Issue
Are others likely hitting this too?
Flaky / Intermittent
Is it intermittent?
Verified / Reproducible
Can you reproduce it reliably?
Loading…
#api#latency issue#model loading#dependency error#configuration error

Still need to ship something?

×6

Another batch ranked right after the header list — different links, same matching logic.

Back to top recommendations

TRENDING