Tweet by fatih

October 17, 2025

I'm going to die on this hill, but this is exactly my thinking. NO ONE SHOULD CARE ABOUT THE MODEL. It's not my job to understand the differences. Do you know which version Google is using for their search, or which version your favorite App is running? No. Why should I care which model my Agent is running? Currently both @AmpCode and @cursor_ai got it correct. Amp doesn't advertise and just calls it "smart", and Cursor comes with "auto", which is basically telling you "just code, don't think about the model, we'll give you the best user experience". So, the agent should be the doing what it is at the best, meaning it should be able to understand my prompt properly, and execute it. Does it run Sonnet 4.5? So what? Who cares. All I care is the output. If I don't like the output of it, I'm free to use other agents.

Author
fatih
Date
October 17, 2025
Open on X