you try to optimize for tokens;
I just cruise on the most fancy model on max settings hitting my budget in 1 prompt.

We are not the same