CLI, App, and IDE: Clean up the model picker
Resolved 💬 3 comments Opened Mar 6, 2026 by miraclebakelaser Closed Mar 8, 2026
💡 Likely answer: A maintainer (github-actions[bot], contributor)
responded on this thread — see the highlighted reply below.
The model picker on all interfaces has become too bloated and confusing. It should be pruned to reduce overlap.
- On Twitter, both @tibo-openai and @ambrosino-oai have said that
gpt-5.4is the model. It should be the default. gpt-5.3-codexandgpt-5.2should have "latest" removed from their descriptions.- Superseded models should be removed from the picker entirely. This includes
gpt-5.2-codex,gpt-5.1-codex-max, andgpt-5.2. There's a case forgpt-5.1-codex-ministicking around because non-pro users don't have access togpt-5.3-codex-spark.
Here's what the model picker should look like:
<img width="826" height="166" alt="Image" src="https://github.com/user-attachments/assets/8104047d-a9f1-4fd3-9680-1555e2455b2c" />
The downsides are minimal because the picker provides clear guidance on how to use legacy models.
3 Comments
Potential duplicates detected. Please review them and close your issue if it is a duplicate.
Powered by Codex Action
Not a duplicate.
Closing, dupe of https://github.com/openai/codex/issues/13736