Gemini 3.1 Pro just went live, and it’s basically a massive brain upgrade that more than doubles the reasoning power of the previous version.
Scored 77.1% on the ARC-AGI-2 benchmark, proving it can solve brand-new logic puzzles it has never seen before rather than just repeating training data.
Massive jump from the 31.1% score on the previous version, meaning it actually "thinks" through problems.
Generates animated SVGs from text which is perfect for our no-code web projects.
Powers the new Antigravity platform for building agents that can navigate live APIs.
Available now with higher limits for Google AI Pro and Ultra users.