OpenAI Codex CLI v0.146.1 Changelog
What changed in Codex CLI v0.146.1, released 2026-08-05 โ explained in plain English.
- Tighter safety rules for AI models that can write computer code.
- The terminal now clearly explains when it changes your permissions.
- This update is a quick maintenance patch to make automated coding safer.
Want a heads-up when OpenAI Codex CLI updates?
We email you the moment a new OpenAI Codex CLI version drops โ plus plain-English release notes like these. No spam.
This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.
This release is a small but important safety update for the OpenAI Codex CLI, which is a tool that lets developers run AI coding assistants right in their computer's terminal. It mostly matters to developers who use highly capable AI models to write or manage code, ensuring the AI double-checks its work before doing anything risky.
Fixes
- Safer AI code reviews: Some AI models are "cyber-capable," meaning they are smart enough to write complex computer code or interact with system files. Previously, the automatic safety review for these models wasn't strict enough by default. This fix makes the AI double-check its actions more carefully before running code, keeping your computer safer.
- Clearer permission warnings: When the tool needs to change what it is allowed to do on your computer (its "permissions"), it used to just do it quietly. Now, it explains exactly what is changing right there in the terminal text interface, so you are never left in the dark about what the AI is doing.
Breaking changes
No breaking changes in this release.
How to upgrade
Update to OpenAI Codex CLI v0.146.1 by running:
codex --upgrade
This is our plain-language summary. Read the complete, official notes on the OpenAI Codex CLI official changelog โ.
