Mobile Codex shows Speed as Standard while desktop shows Fast for the same GPT-5.5 Extra High session

Summary

The Codex mobile UI and the Codex desktop UI appear to show inconsistent Speed settings for the same GPT-5.5 Extra High Codex session.

On desktop, the session is shown as using GPT-5.5 with Extra High intelligence and Fast speed. On mobile, opening the same Codex session and tapping the model/settings menu shows:

• Model: GPT-5.5
• Intelligence: Extra High
• Speed: Standard

Both the mobile app and the desktop Codex app have been updated to the latest available versions.

Environment

Desktop:

• App: Codex desktop app
• Version: 26.513.20950 (2816)
• Model shown in the composer: GPT-5.5 Extra High
• Speed shown on desktop: Fast

Mobile:

• App: Codex mobile UI / mobile app
• Device: iPhone
• Model menu shows:
• Model: GPT-5.5
• Intelligence: Extra High
• Speed: Standard
• Mobile app is updated to the latest available version as of the report

Actual behavior

The same Codex session appears to report different Speed values depending on the client:

• Desktop: Fast
• Mobile: Standard

The model and intelligence level appear to match, but the Speed value does not.

Expected behavior

The Speed setting should be consistent across clients for the same session.

If the actual backend setting is Fast, the mobile UI should also show Fast. If the actual backend setting is Standard, the desktop UI should not show Fast. If the mobile client does not support displaying or changing Speed accurately, it should not silently show a conflicting value.

Steps to reproduce

1. Open a Codex session on the desktop Codex app.
2. Select GPT-5.5 with Extra High intelligence and Fast speed.
3. Open the same Codex session from the mobile Codex UI / mobile app.
4. Tap the model/settings menu on mobile.
5. Observe that Model and Intelligence match the desktop selection, but Speed displays Standard instead of Fast.

Impact

This is confusing because users cannot tell which Speed setting is actually active. It also makes it difficult to understand model behavior, expected latency, and usage behavior across devices. Users may think they are running the same configuration across desktop and mobile, while the UI reports different Speed settings.

Suggested fix

Please make the Speed value synchronized and displayed consistently between Codex desktop and mobile clients, or clarify in the UI if mobile is showing a fallback/default display value rather than the actual active session configuration.

<img width="941" height="2048" alt="Image" src="https://github.com/user-attachments/assets/b9641436-6e6d-4652-a5bb-eb6535dbffb5" />

<img width="1812" height="958" alt="Image" src="https://github.com/user-attachments/assets/fbcff803-7979-49c8-8e31-0abf6fccde7e" />