← Back to context

Comment by jdthedisciple

1 year ago

I heard Claude Sonnet was supposed to be "so much better" than GPT4o at coding.

Recently I had both of them port an entire scraping script from python to C#.

Surprisingly, Claude's version was borderline unusable whereas GPT4o's ran perfectly as expected right away (besides forgetting to tell me about a nuget package at the first shot)!

So yea, was a bit disappointed in Claude given the hype about it that's all. Would be curious about other people's experiences. I still personally find GPT4o to be the most intelligent model to date.