I try not to shill or be unfair to any particular AI company but jesus christ using a $200 sub for Claude and chatgpt side by side inside the same project shows how glaringly obvious the gap is.
Codex is so slow, doesnt complete tasks fully, meanwhile at any moment I can give claude code control over codex's mess, give it my expectations and IT GETS DONE.
I try to be fair, but this is ridiculous, I will almost certainly be swapping to 2 claude subs instead and maybe still retain a pro100 for testing but christ almighty if claude wasnt so terrible at marketing openai would be hemorrhaging subscribers if most people realized how massive the gap actually is.