Incorrect model in the subagent's chat UI

[Artemonim]

Where does the bug appear (feature/product)?

Cursor IDE

Describe the Bug

See screenshots

Steps to Reproduce

Ask Opus to work through subagents

Screenshots / Screen Recordings

Operating System

Windows 10/11

Version Information

Version: 3.2.0-pre.35.patch.0 (system setup)
VSCode Version: 1.105.1
Commit: d1a36a628792b1e615917859916162ea8a2f7e10
Date: 2026-04-17T01:50:29.007Z
Layout: editor
Build Type: Stable
Release Track: Nightly
Electron: 39.8.1
Chromium: 142.0.7444.265
Node.js: 22.22.1
V8: 14.2.231.22-electron.0
OS: Windows_NT x64 10.0.22631

For AI issues: which model did you use?

Opus 4.7 Thinking Max, GPT-5.3-Codex-XHigh

For AI issues: add Request ID with privacy disabled

Privacy on, MAX mode on:
agent: 438eb150-839b-4101-ab53-6906c2fd2f44
sub: 882edf6f-bbc8-4551-bf72-5da930efa53f

Does this stop you from using Cursor

Sometimes - I can sometimes use Cursor

[Artemonim]

[mohitjain (Mohit)]

Hey @Artemonim,
Thanks for reporting this again. I can see from your screenshots that the usage log correctly shows gpt-5.3-codex-xhigh was used for the subagent, but the chat header displays “GPT-5.4” – so the correct model did run, and this is a display-only issue.

I wasn’t able to reproduce this on my end (the subagent header showed the correct model name), so this may be specific to the Nightly build you’re on. Could you check if this still happens after updating to a newer Nightly? If it persists, a short screen recording showing the mismatch would help our team track it down.