Google launches Gemini 3.1 Pro scoring 77.1% on ARC-AGI-2
Technology
Google just rolled out Gemini 3.1 Pro, its newest artificial intelligence (AI) model with a big boost in reasoning skills, scoring 77.1% on a tough logic benchmark (ARC-AGI-2).
That's a huge jump from last year's Gemini 3 and shows Google is stepping up its game in making AI smarter and more useful.
Gemini 3.1 Pro available across Google
Gemini 3.1 Pro isn't just for techies, it's available to developers through Google's AI Studio, to businesses via Vertex AI, and even to everyday users on NotebookLM and the Gemini app.
This launch highlights how Google is focused on building AIs that actually help out in real-world situations, not just show off in labs.