You’re not ready for how visual Google just became
Google has released Gemini 3, its most advanced model so far, and it’s now built directly into Google Search through AI Mode.
This gives Search better reasoning, a stronger grasp of complex questions, and new interactive visuals that adjust to what you’re asking.
Gemini 3 Pro is available first to U.S. Google AI Pro and Ultra subscribers under the “Thinking” option, with wider access coming soon.
Because the model understands intent more accurately, Search can scan the web more effectively and surface more relevant, credible information.
Google will also start using Gemini 3 automatically for harder questions, while faster models handle simple ones.
The update also introduces a new generative interface. Instead of plain text answers, Gemini 3 can create visual layouts, tables, images, and even custom tools.
Three things to know:
Gemini 3 boosts Search with stronger reasoning and deeper understanding.
AI Mode now generates interactive visuals, tools, and simulations.
U.S. subscribers get first access, with broader rollout planned.
Visuals you didn’t ask for but need
It can build real-time simulations, like showing how the three-body problem works, or generate calculators on the spot for topics like mortgages.
All responses still point to high-quality sources if you want to keep reading.
Google says more improvements are on the way as people start using these new features.
Poor Altman thought he was going to have a quiet week… - MV