I follow things pretty closely, and I was genuinely unsure if GPT5-pro was a real thing or just equivalent to setting GPT-5 to high reasoning.
Does the fact it's entering the API confirm that it's a fully separate thing which is not widely benchmarked, and which was not part of the GPT-5 stream?
This particular thing has not been well communicated AFAICT.
I live-blogged the keynote here, adding a little bit of extra commentary and links along the way: https://simonwillison.net/2025/Oct/6/openai-devday-live-blog...
"Sam says today Codex is out of preview and into GA. But which one! Codex CLI or Codex Cloud or both?"
The cloud version as "Starting October 20, Codex cloud tasks will also begin counting towards your Codex usage"
I follow things pretty closely, and I was genuinely unsure if GPT5-pro was a real thing or just equivalent to setting GPT-5 to high reasoning.
Does the fact it's entering the API confirm that it's a fully separate thing which is not widely benchmarked, and which was not part of the GPT-5 stream?
This particular thing has not been well communicated AFAICT.
GPT-5 Pro is not just GPT-5 with higher reasoning effort - it’s a different thing.
Under the hood, it’s based on GPT-5, and you’ll notice all of GPT-5’s personality traits in its outputs, so it still bears the GPT-5 brand.
The "live" coding demo felt very staged with codex reasoning set at "low" and everything working ootb.
Probably didn't want to risk the same embarrassment that was the Meta live demo.
Every single person I know had nothing but respect for the meta people doing it live. I think they garnered more attention that way
I created transcript for it for anyone wanting to read instead of listen!
https://videotobe.com/play/youtube/hS1YqcewH0c