I tried it in an 8h coding session. It performs worse than Claude 4 for me, and it’s slower. It made me wait for 10 minutes at a time. Eventually, I gave up, used my brain to understand the problem properly, and hand held Claude to a solution, which took about 1h. I think I lost most of my time with GPT-5 in loops where it fixes one thing at the expense of another while the general approach looked sufficiently sound to fool me but not sufficiently sound to eventually work in all cases.
This might be due to high traffic and not because the model is worse. I’ll give it another chance when the hype fades.
Veselin.blog 
I’m sad to hear it didn’t work well for you, but I’m even happier that you were able to solve the problem by other means. I tried the new model as well when it came out and I liked it. Did not use it for code of course, since I’m not in that field, but I used it for other things and I was pretty amazed at the results and how funny it was.
LikeLike
Oh, these models are all incredible, I agree. Just that, I don’t see GPT-5 as superior to Claude 4. In my case, I lost about 8h, and the gain from these 8h was that I was absolutely sure that what it tried to convince me to do is wrong.
With Claude, I tried giving similar prompts, similar context, similar directions, we got in a direction that worked and made it work within 1-2h.
LikeLiked by 1 person