By skyforbes I’m a heavy Gemini Pro 2.5 thinking user with Google grounded search via vertex AI. It’s my go to model from nearly everything and it’s been for the last six months. very few niche use cases I’m doing with Claude models.
Not my question has anybody seen improvement of 3 versus 2.5 pro for writing and research usecases (not coding)? 2.5 Pro is top notch for me already and I haven’t seen any, if at all a bit of a regression.