Hacker News Reader

Interesting, I've had the complete opposite experience. Opus 4.1 feels like a generational improvement compared to GPT-5.

It is funny how it can be like this sometimes. I think a lot depends on coding styles, languages, prompting, etc.