Krux
February 20, 2026
Google's Gemini 3.1 Pro Doubles Reasoning Score, Goes Live
Published: February 20, 2026 at 12:56 AM
Updated: February 20, 2026 at 12:56 AM
100-word summary
Google just unleashed Gemini 3.1 Pro, its supercharged reasoning model now rolling out across consumer apps, developer tools, and enterprise platforms. The upgrade more than doubles its predecessor's performance, hitting a 77.1% score on the tough ARC-AGI-2 benchmark. Think better problem-solving for complex tasks like turning data into interactive dashboards, creating animated website graphics from text prompts, or designing 3D visualizations. It's accessible via Gemini API, Vertex AI, the Gemini app, and NotebookLM. Google's keeping it in preview mode for now, testing ambitious AI agent workflows before the full launch. This positions Google to compete harder in the enterprise AI race against rivals.
What happened
Google just unleashed Gemini 3.1 Pro, its supercharged reasoning model now rolling out across consumer apps, developer tools, and enterprise platforms. The upgrade more than doubles its predecessor's performance, hitting a 77.1% score on the tough ARC-AGI-2 benchmark. Think better problem-solving for complex tasks like turning data into interactive dashboards, creating animated website graphics from text prompts, or designing 3D visualizations. It's accessible via Gemini API, Vertex AI, the Gemini app, and NotebookLM. Google's keeping it in preview mode for now, testing ambitious AI agent workflows before the full launch.
Why it matters
This positions Google to compete harder in the enterprise AI race against rivals.