Which model should I use?

Discussion
1

I’ve been hearing recently that the Claude 3.7 Sonnet model has poor performance. I can’t decide which model I should use. Please help me.

2

Everyone has a different opinion about models, some people will think they are terrible and others will think they work great. You can try them all and find what works best for you.

1 Like
3

This is fluctuating. All models act different each day. Don’t expect any consistency. Take it day by day. Keep chat short. Be specific in your prompt. If it can’t fix it in less then 5 tries, start over with another model, or else be victimized with only token spam. That’s just the way it is today :person_shrugging:

4

It depends on your needs. For me:

Frontend: Claude 3.7
Backend: Gemini 2.5
Boilerplate: Claude 3.5

Or if it returns a lot of errors use Auto.

5

For me:
Gpt-4o: for simple tasks and getting suggestions on how to possibly structure things, etc.
Claude-3.5-sonnet: for general coding
Claude-3.7-sonnet: for UI/web stuff. It’s great at css and next.js