What's your AI model ranking so far?

Discussion
1

Cursor provides 500 fast requests every month for Pro users.

It always run out very quick, so sometimes has to choose another models, like GPT 4o-mini for quick response.

What’s your AI model ranking so far ?
Ask for suggestion, so I can choose when run out of fast requests.

For me:

Premium models:

  1. Claude-3.5-Sonnet
  2. Claude-3.7-Sonnet
  • Supposedly 3.7 should be smarter, but after I tried couple of times, it doesn’t work better than 3.5

Normal models (quick request and unlimited):
3. GPT 4o-mini
4. DeepSeek R1
5. (don’t know)

But basically, Claude-3.5.Sonnet is my best and the only choice, the other like 4o-mini, DeepSeek is not that good. But can handle some easy tasks, or answer some easy questions.

What’s your ranking ?

2

could you please elaborate why you ranking sonnet 3.5 higher than 3.7?