Skip to content

Model changes to Medium uninstructed #11013

Closed
Closed
Model changes to Medium uninstructed#11013
Labels
CLIIssues related to the Codex CLIbugSomething isn't working

Description

@Redstar1337
Redstar1337
opened on Feb 7, 2026

What version of Codex CLI is running?

0.98

What subscription do you have?

Pro

Which model were you using?

5.3 Codex

What platform is your computer?

PS7 Win10

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

PS7

What issue are you seeing?

Worked for 2m 59s ──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
• I found a potential regression risk in the patch and want to verify it empirically with a short replay: internal netting fills are now written to blotter, which may unintentionally affect turnover/fee
accounting. I’ll run a short-window smoke replay and inspect blotter plus turnover consistency.

/status

╭─────────────────────────────────────────────────────────────────────────────────╮
│ >_ OpenAI Codex (v0.98.0) │
│ │
│ Visit https://chatgpt.com/codex/settings/usage for up-to-date │
│ │
│ Model: gpt-5.3-codex (reasoning xhigh, summaries auto) │
│ Directory: E:\Python 3.12 │
│ Approval: never │
│ Sandbox: danger-full-access │
│ Agents.md: AGENTS.md │
│ Account: xxx (Pro) │
│ Collaboration mode: Default │
│ Session: 019c371b-daab-7ec3-9f22-10ba8c2c4738 │
│ Context window: 29% left (187K used / 258K) │
│ 5h limit: [███████████████████░] 95% left (resets 20:48) │
│ Weekly limit: [██████████░░░░░░░░░░] 49% left (resets 01:13 on 10 Feb) │
╰─────────────────────────────────────────────────────────────────────────────────╯

› Yes proceed

/status

╭─────────────────────────────────────────────────────────────────────────────────╮
│ >_ OpenAI Codex (v0.98.0) │
│ │
│ Visit https://chatgpt.com/codex/settings/usage for up-to-date │
│ information on rate limits and credits │
│ │
│ Model: gpt-5.3-codex (reasoning medium, summaries auto) │
│ Directory: E:\Python 3.12 │
│ Approval: never │
│ Sandbox: danger-full-access │
│ Agents.md: AGENTS.md │
│ Account: (Pro) │
│ Collaboration mode: Plan │
│ Session: 019c371b-daab-7ec3-9f22-10ba8c2c4738 │
│ │
│ Context window: 21% left (207K used / 258K) │
│ 5h limit: [███████████████████░] 94% left (resets 20:48) │
│ Weekly limit: [██████████░░░░░░░░░░] 48% left (resets 01:13 on 10 Feb) │
╰─────────────────────────────────────────────────────────────────────────────────╯

What steps can reproduce the bug?

I'm not sure yet, possibly turning on Plan mode causes it to happen but needs further testing to reproduce.

What is the expected behavior?

No response

Additional information

No response

Activity

  1. Redstar1337
    added
    bugSomething isn't working
    on Feb 7, 2026
  2. github-actions
    added
    CLIIssues related to the Codex CLI
    windows-osIssues related to Codex on Windows systems
    model-behaviorIssues related to behaviors exhibited by the model
    on Feb 7, 2026
  3. etraut-openai
    removed
    model-behaviorIssues related to behaviors exhibited by the model
    on Feb 7, 2026

  4. etraut-openai commented on Feb 7, 2026

    @etraut-openai
    etraut-openai
    on Feb 7, 2026
    Contributor

    Thanks for the bug report. Let us know if you come up with reliable repro steps.

  5. Redstar1337 commented on Feb 7, 2026

    @Redstar1337
    Redstar1337
    on Feb 7, 2026
    Author

    I'm almost certain entering plan mode switches it to the default medium model, happened more than once now switching to plan mode.

  6. fullofcaffeine commented on Feb 12, 2026

    @fullofcaffeine
    fullofcaffeine
    on Feb 12, 2026

    +1, having the same issue. I notice because it becomes too fast and then I realize it's a bit shallow on the replies. Annoying because plan mode is quite useful.

  7. fullofcaffeine commented on Feb 12, 2026

    @fullofcaffeine
    fullofcaffeine
    on Feb 12, 2026

    Yep, switching to the plan mode does it. Codex 0.98.0, MacOS 15.4 (24E248).

    1. Fire a new session, send the /model. command and make sure mode is something other then 'gpt5.3-codex-medium', i.e 'GPT5.3-codex-high';
    2. Press shift+tab to switch to the plan mode;
    3. Send the /model command again and you'll see gpt5.3-codex-medium selected again 😞
  8. etraut-openai
    marked Plan mode overrides explicitly selected high/xhigh effort to medium #11962 as a duplicate of this issue on Feb 17, 2026

  9. guidedways commented on Feb 17, 2026

    @guidedways
    guidedways
    on Feb 17, 2026
    Contributor

    Same: #11979

  10. etraut-openai
    removed
    windows-osIssues related to Codex on Windows systems
    on Feb 17, 2026
  11. etraut-openai
    marked gpt-5.3-codex xhigh → auto switches to gpt-5.3-codex medium mid way #11979 as a duplicate of this issue on Feb 17, 2026
  12. etraut-openai
    marked Codex CLI resets model_reasoning_effort to Medium (with proof) #12042 as a duplicate of this issue on Feb 17, 2026
  13. etraut-openai
    marked Choose gpt5.2 xhigh maually but it sometimes change to medium during the execution #12138 as a duplicate of this issue on Feb 18, 2026

  14. 20 remaining items

  15. saidelike commented on Mar 4, 2026

    @saidelike
    saidelike
    on Mar 4, 2026

    @etraut-openai this is NOT patched i.e. this does not seem to be fixed. Even in 0.107.

  16. etraut-openai commented on Mar 4, 2026

    @etraut-openai
    etraut-openai
    on Mar 4, 2026
    Contributor

    @saidelike, can you clarify what behavior you're seeing?

  17. am-will commented on Mar 4, 2026

    @am-will
    am-will
    on Mar 4, 2026

    You have to configure your model reasoning level defaults if you want them to change.

    At the top of your config file, you can add something like this. Cheers!

    model = "gpt-5.3-codex"
model_reasoning_effort = "medium"
plan_mode_reasoning_effort = "high"

  18. saidelike commented on Mar 4, 2026

    @saidelike
    saidelike
    on Mar 4, 2026

    @saidelike, can you clarify what behavior you're seeing?

    By default, it automatically changes the reasoning effort to medium when switching to /plan mode:

    I start codex, it uses xhigh level:

     % codex          
╭───────────────────────────────────────────────────╮
│ >_ OpenAI Codex (v0.106.0)                        │
│                                                   │
│ model:     gpt-5.3-codex xhigh   /model to change │
...

    I switch to plan mode, I see it changes to medium:

    /plan

• Model changed to gpt-5.3-codex medium for Plan mode.

    Expected: we expect codex to not change the level when switching to plan mode, by default.

  19. am-will commented on Mar 4, 2026

    @am-will
    am-will
    on Mar 4, 2026

    @saidelike, can you clarify what behavior you're seeing?

    By default, it automatically changes the reasoning effort to medium when switching to /plan mode:

    I start codex, it uses xhigh level:

     % codex          
╭───────────────────────────────────────────────────╮
│ >_ OpenAI Codex (v0.106.0)                        │
│                                                   │
│ model:     gpt-5.3-codex xhigh   /model to change │
...

    I switch to plan mode, I see it changes to medium:

    /plan

• Model changed to gpt-5.3-codex medium for Plan mode.

    Expected: we expect codex to not change the level when switching to plan mode, by default.

    You need to configure your plan mode reasoning level in your config file. When you switch to plan mode and change reasoning levels, it will ask you if you want to save it as a default.

  20. saidelike commented on Mar 5, 2026

    @saidelike
    saidelike
    on Mar 5, 2026

    You need to configure your plan mode reasoning level in your config file. When you switch to plan mode and change reasoning levels, it will ask you if you want to save it as a default.

    Thanks for this, it indeed works. However, without setting this in the config file, we would expect the level to NOT change when switching to /plan mode, as described above.

  21. reverofevil commented on Apr 6, 2026

    @reverofevil
    reverofevil
    on Apr 6, 2026

    Switching from xhigh to medium to do the thinking honestly doesn't sound very logical to me, as well as having to plan_mode_reasoning_effort = "high" in the config file instead of switching it through /model.

    @etraut-openai Do we need to open a separate bug for this, or is this considered a feature request because it never worked properly?

  22. kendonB commented on May 4, 2026

    @kendonB
    kendonB
    on May 4, 2026

    @etraut-openai this should be unpinned?

  23. etraut-openai
    unpinned this issue on May 4, 2026
  24. monkeycode-ai
    added a commit that references this issue on Jun 8, 2026

    Improve Plan mode reasoning selection flow (#12303)

    1e22eba
  25. wangjiecloud
    added a commit that references this issue on Jun 27, 2026

    Improve Plan mode reasoning selection flow (openai#12303)

    e1e30c3

  26. yevon commented on Jul 27, 2026

    @yevon
    yevon
    on Jul 27, 2026

    This bug is happening again each time you start a new chat with 5.6 sol and both in macos and windows, this should be fixed i'm wasting tons of tokens with this

  27. shafqatevo
    added a commit that references this issue on Aug 7, 2026

    Improve Plan mode reasoning selection flow (#12303)

    dd659a5
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    CLIIssues related to the Codex CLIbugSomething isn't working

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions