macOS Codex: gpt-5.5 returns 404 Model not found while gpt-5.4 works

Resolved 💬 7 comments Opened Jun 8, 2026 by mayuhang2008-byte Closed Jun 12, 2026
💡 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)?

26.602.40724

What subscription do you have?

Plus

What platform is your computer?

Darwin 22.3.0 arm64 arm

What issue are you seeing?

When using Codex on macOS with ChatGPT authentication, gpt-5.5 is listed locally as an available model but real requests fail server-side.

Observed errors:

  • unexpected status 404 Not Found: Model not found gpt-5.5-codex-1p-exp-p-0604-811f2f-ev3-text-1-treatment
  • unexpected status 404 Not Found: Model not found gpt-5.5

Working model:

  • gpt-5.4

Failing model:

  • gpt-5.5

This happens in the Codex app and is also reproducible from the bundled CLI while using the same ChatGPT login.

What steps can reproduce the bug?

  1. Log into Codex with ChatGPT authentication.
  2. Confirm the local model catalog lists gpt-5.5.
  3. Run:

codex exec --skip-git-repo-check -m gpt-5.5 'Reply with exactly: OK'

  1. Observe websocket prewarm failure for the treatment model:

gpt-5.5-codex-1p-exp-p-0604-811f2f-ev3-text-1-treatment

  1. Observe HTTP fallback and final 404 for gpt-5.5.
  2. Run:

codex exec --skip-git-repo-check -m gpt-5.4 'Reply with exactly: OK'

  1. Observe that gpt-5.4 works normally.

Also reproduced with:

  • codex exec --skip-git-repo-check -c 'model_reasoning_effort="low"' -m gpt-5.5 'Reply with exactly: OK'
  • codex exec --skip-git-repo-check -c 'model_reasoning_effort="medium"' -m gpt-5.5 'Reply with exactly: OK'
  • codex exec --skip-git-repo-check -c 'model_reasoning_effort="high"' -m gpt-5.5 'Reply with exactly: OK'
  • codex exec --skip-git-repo-check -m gpt-5.5-2026-04-23 'Reply with exactly: OK' which returns: The 'gpt-5.5-2026-04-23' model is not supported when using Codex with a ChatGPT account.

Request IDs captured:

  • 7463da3a-4876-4a94-9015-c47e07e0d2b6
  • acf1e346-e322-4f72-8a19-7770b2897d06
  • be50d67b-cc94-494e-90e3-028989eb1bc9
  • b247b419-dd98-4892-a392-2ec1877ba820

What is the expected behavior?

If gpt-5.5 is listed as available locally for a ChatGPT Plus account in Codex, selecting or invoking gpt-5.5 should work normally instead of failing with a 404 model-not-found error.

Additional information

Environment:

  • macOS 13.2.1 (Apple Silicon)
  • Codex App 26.602.40724
  • Bundled CLI 0.137.0-alpha.4
  • Auth mode: ChatGPT

What was ruled out:

  • Old client version
  • Local default model misconfiguration
  • Thread-specific state
  • Reasoning-effort-specific failure

This looks like a ChatGPT-auth entitlement or backend routing mismatch for gpt-5.5 in Codex on this account.

View original on GitHub ↗

7 Comments

github-actions[bot] contributor · 1 month ago

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

  • #27010
  • #26927
  • #26916
  • #26910
  • #26892

Powered by Codex Action

gabriel199002 · 1 month ago

Hi
Same issue on Windows 11 with Codex Desktop v26.602.40724.

Environment:

  • Windows 11
  • ChatGPT Plus
  • Codex Desktop 26.602.40724
  • Auth mode: ChatGPT

Observed behavior:
All requests fail immediately with:

unexpected status 404 Not Found: Model not found gpt-5.5
url: https://chatgpt.com/backend-api/codex/responses

This occurs even for simple prompts such as:
"Explain Python in 5 lines"

Troubleshooting already performed:

  • Restarted Codex
  • Logged out and back in
  • Regenerated %LOCALAPPDATA%\OpenAI\Codex
  • Renamed and regenerated %USERPROFILE%.codex
  • Created a clean profile and new workspace

The issue persists.

Feedback ID:
no-active-thread-019ea830-54c0-7741-ad13-172690d2c098

talescasalta · 1 month ago

Same here
Feebackid 019ea837-17fa-7033-abe3-88e296ef3d81

LaFerrari13 · 1 month ago

same, horrifying

GGBondBlueWhale · 1 month ago

Same problem here

ax-openai · 1 month ago

Thanks for your patience. The backend issue causing gpt-5.5 requests to return 404 Model not found is now resolved.

Please try again. If it still fails, share a fresh request ID and timestamp.

ax-openai · 1 month ago

Closing as resolved. If this recurs, please open a new issue with a fresh request ID and timestamp.