Expose thinking level param for Codex CLI
Resolved 💬 4 comments Opened Jan 16, 2026 by louenas Closed Mar 20, 2026
💡 Likely answer: A maintainer (etraut-openai, contributor)
responded on this thread — see the highlighted reply below.
What feature would you like to see?
Is your feature request related to a problem?
cannot set the reasoning level when calling Codex CLI from shell script or from terminal. Other coding agents postfix the model name with the thinking level but Codex has no parameter neither other simple way to do the same. I know it's possible to customize the launch of Codex using profiles but it become complicated when trying to integrate Codex in a sort of Ralph loop. For example I would like to create the PRD and Backlog using high and code using medium.
Describe the solution you'd like
expose the thinking param
Additional context
nothing
Additional information
_No response_
4 Comments
Codex supports setting the reasoning effort via config overrides. You can set the
model_reasoning_efforteither inconfig.tomlor as a one‑off CLI override with--config/-c. Here's an example:super, thanks for the quick feedback :) I'll try it
isn't it possible to change it within the chat session using something like
/effort?@agn-7, after you choose a model with the
/modelsetting you'll be given a prompt to set the effort on that model.