n way to set new quick chat deafult model and reasoning.

Open 💬 0 comments Opened Jul 14, 2026 by serdox

What variant of Codex are you using?

app

What feature would you like to see?

New Quick Chats initialize with thinkingEffort: null, which makes them fall back to the server’s low/default effort. Your Codex setting model_reasoning_effort = "high" only affects Codex tasks, not ordinary ChatGPT Quick Chat.

Additional information

_No response_

View original on GitHub ↗