I’m writing this as a friendly PSA for fellow devs who use Codex (especially via codex CLI) and are tempted by the $200 ChatGPT Pro upgrade.
Because I fell into a trap.
What I expected when I upgraded
I was on the $20 plan, and I upgraded to Pro thinking:
- “Nice, now Codex CLI will use GPT-5.2 Pro.”
- “And GitHub PR reviews via Codex will clearly be the Pro model too.”
That felt like a reasonable assumption: Pro = best model, so Codex = best model, right?
All the above was advised by ChatGPT..
What actually happened
After upgrading, I opened Codex CLI on my Mac…
…and I did not see a “GPT-5.2 Pro” model option.
What I saw was basically: Codex models / GPT-5.2 variants, but not a clear “Pro” switch the way you get in the ChatGPT app/web model picker.
And for the GitHub @codex review flow, there isn’t an obvious indicator in the PR that says “this review used GPT-5.2 Pro” either.
So if you’re upgrading specifically because you think you’ll get a brand new “Pro Codex model” in the CLI… you might be disappointed.
Upon clarifying with ChatGPT PRO, I was replied with:
Sorry… I previously implied that, and that was wrong.
So what does Pro actually buy you (for Codex)?
From what’s visible as a user: more capacity and priority — higher limits, more cloud tasks, more code reviews.
That’s legit value if you’re hitting limits or running Codex heavily.
But it’s not the same as:
“Upgrading to Pro unlocks a new Codex model you can select.”
My take
If you’re not hitting limits on the $20 plan, don’t assume the $200 upgrade magically turns Codex into a different beast.
Hope this saves someone else $180 and some frustration.
And don’t blindly trust GPT answers..