21
Warning: ChatGPT 4o vs DeepSeek R1 on a coding task was not even close
I spent last weekend testing both on the same task: building a simple booking system in Python. DeepSeek R1 got the whole thing working on the first try, while ChatGPT 4o left out half the error handling and needed 4 follow-ups to fix it. Has anyone else seen this big of a gap between the two for real projects?
2 comments
Log in to join the discussion
Log In2 Comments
martinez.anna8d ago
oh man, I feel this so hard it's almost PAINFUL because I SUCK at coding and rely on these tools to look like I know what I'm doing. last week I asked ChatGPT to write a simple script to rename a bunch of files and it gave me something that looked right but then deleted my entire downloads folder instead. not even kidding. DeepSeek has been more reliable for me too, especially on the first try. I think it's because DeepSeek seems to actually THINK through the edge cases while ChatGPT just spits out the most common path and hopes for the best. my advice? test both on a dummy folder first before unleashing them on anything important, believe me I learned that one the hard way.
3