/model does not list GPT-5.6 models that are available via -m

Open 💬 5 comments Opened Jul 9, 2026 by ikhristich

What version of Codex CLI is running?

Codex CLI: v0.143.0

What subscription do you have?

Plus

Which model were you using?

gpt-5.6

What platform is your computer?

macOS

What terminal emulator and version are you using (if applicable)?

iTerm2

Codex doctor report

What issue are you seeing?

Description

The Codex CLI can successfully start GPT-5.6 models when they are specified with the -m flag, but the interactive /model picker does not display them.

This makes it appear that GPT-5.6 models are unavailable, even though they are fully usable.

Environment

  • Codex CLI: v0.143.0
  • macOS
  • ChatGPT Plus account
  • Logged in with ChatGPT account (not API key)

What steps can reproduce the bug?

Steps to reproduce

  1. Start Codex with any GPT-5.6 model:

codex -m gpt-5.6-sol

or

codex -m gpt-5.6-terra

or

codex -m gpt-5.6-luna

  1. Verify that the session starts successfully and the header displays, for example:

model: gpt-5.6-sol medium

  1. Run:

/model

Actual behavior

The model picker only shows:

  • gpt-5.5
  • gpt-5.4
  • gpt-5.4-mini

GPT-5.6 models are missing, even though the current session is already running on GPT-5.6.

What is the expected behavior?

Expected behavior

The /model picker should list every model available to the authenticated user, including:

  • gpt-5.6-sol
  • gpt-5.6-terra
  • gpt-5.6-luna

or at minimum display the currently active model.

Additional information

The same ChatGPT account has access to GPT-5.6 models in the Codex App.

The CLI can also successfully start GPT-5.6 models when they are explicitly specified using the -m option.

This appears to be a UI/model picker issue rather than an authorization or model availability problem.

View original on GitHub ↗

This issue has 5 comments on GitHub. Read the full discussion on GitHub ↗