Add the thinking tokens back to the Codex transcript
Resolved 💬 4 comments Opened Nov 28, 2025 by chadlangston Closed Mar 29, 2026
💡 Likely answer: A maintainer (etraut-openai, contributor)
responded on this thread — see the highlighted reply below.
What feature would you like to see?
Being able to review the thinking helps me guide the model and saves me from wasting precious time prompting the model over and over again to do tasks that it is no where close to achieving.
Additional information
I still see a few sparse thinking tokens in the transcript, but no where near the amount of reasoning I used to be able to see.
4 Comments
Are you using the CLI or IDE extension? Does https://github.com/openai/codex/issues/5339 cover your feature request? If so, let's close this as a duplicate.
Also, if you're using the CLI, try pressing ctrl+T to toggle between terse mode and full mode.
I am using the CLI. The CTRL+T transcript used to show all the thinking tokens. With each new release there have been fewer and fewer thinking tokens. Now there are almost none. I think it's helpful to see the thinking tokens.
This feature request hasn't received enough upvotes, so I'm going to close.