Holy crap, rumor has it that Gemini 3.5 Pro's internal test scores have surpassed Fable 5 and GPT-5.6? In Google's private evaluations, Gemini 3.5 Pro's metrics completely crushed the yet-to-be-released Claude Fable 5 and GPT-5.6. Compared to the previous 3.1 Pro, the new model has achieved a groundbreaking leap in zero-shot performance without any prompt guidance. The model is currently undergoing final private validation, and all preparations for public release are ready. The internal launch date is locked in for July 17. I was so shocked I immediately went on Xianyu to top up a few accounts