How do y track performance of gpt-4o

I’m working, starting with gpt-4o version…
After a while, got the alert:
You’ve reached the current usage cap for GPT-4. You can continue with the default model now, or try again after 3:46 PM

But how can I trully appreciate the performance of gpt-4o, if AI jumps automatically, without even noticing, to gpt-4 ? I’m using a Paid account.

What’s your experience with this issue ?