Control over which model to use for terminal generations

Feature Requests
1

I just tried the terminal command generation feature and it cost me 1 fast premium request. I feel that for simple tasks like terminal command generation Cursor shouldn’t use precious premium requests.

Would be nice to have some control over this.