Regarding the use of the new pricing model

Discussions
1

I’ve noticed a difference when testing the old and new Claude-4 models on Cursor. The new Claude-4 model seems a bit smarter. Is this just my imagination?