for everyone's sanity there should be a place you can see the actual underlying model used for a given chat with ChatGPT. there's basically no reasonable way for any of us to discuss and compare results without it. i don't even know if "pro" is multi-model or or not.