|
|
|
|
|
by plasticbugs
477 days ago
|
|
It definitely depends on what language you’re using and the complexity. But Claude 3.5 and now 3.7 are far superior in my experience for coding tasks. It’s like the difference between ChatGPT 3.5 and 4. It’s far and away more useful and less error-prone for my use cases than other sota models. Cursor + Claude 3.7 in Agent mode is my goto. |
|