Claude 4.1 opus MAX beats gpt-5 no freaking doubt about it

Discussions
1

I was trying to solve a problem with gpt-5 for hours and just one prompt with claude 4.1 opus max solved it. GPT-5 still ■■■■■.

1 Like
2

wait, I was like you

but to be fair it is hard to say which one of GPT-5 or Cursor was trash … unless you clarify whether yes-or-no Cursor had truncated GPT-5 context of the initial prompt even in MAX MODE

3

i also find this gpt5 propaganda being weird, the model is mid at best based on my findings, just being honest

4

Let me tell you something:
You might be right, but it costs several times more and that’s also a factor.
I had a different experience.
These statements like “model X is better than model Y” seem strange to me.
All these models have strengths and weaknesses.
People don’t always judge based on what the model can do, but on what they themselves know how to do there are models that are better, but you might be less skilled at writing them a question or request.

Also, you’re making a claim about which is better so let’s hear it:
Which programming language are you using?
What’s the edge case you need?

5

Of course it does, you’re comparing MAX with a routed model. Compare MAX with MAX. GPT5 PRO MAX crushes Claude 4.1 Opus MAX. And to prove my point, get them both to write a detailed report on any of your current projects and come up with suggestions.

Then, get Claude to read the GPT5 report and see what happens…

6

I am talking about MAX vs MAX