Gemini 3.1 Pro: Advanced Reasoning and Intelligence at Scale
Gemini 3.1 Pro: Advanced Reasoning and Intelligence at Scale
We are excited to announce that Gemini 3.1 Pro is here, and we are officially shipping it at scale. This update represents a significant leap forward in core intelligence and advanced reasoning, specifically engineered for those intricate workflows where a simple answer just isn't enough.
Whether you are a researcher, a creative professional, or a developer, Gemini 3.1 Pro is designed to help you visualize complex topics, organize scattered data, and bring ambitious projects to life.
Beyond Basic Answers: True Reasoning
The real standout of this release is its performance on ARC-AGI-2. This benchmark is unique because it specifically tests for novel logic patterns—the ability to solve problems the AI hasn't seen before—rather than just recalling training data.
On this rigorous test, Gemini 3.1 Pro more than doubles the score of the previous 3 Pro model, marking a massive milestone in the journey toward machine intelligence that can think and adapt like a human.
Where Can You Use It?
We are making Gemini 3.1 Pro available across our ecosystem today:
The Gemini App: Rolling out now for all users to enhance daily productivity.
NotebookLM: Available exclusively to Google AI Pro and Ultra users, providing a powerhouse tool for document synthesis and deep research.
Google AI Studio & Google Antigravity: Developers can access the model in preview via the API to start building the next generation of agentic applications.
Get Started Today
Ready to see what the next level of AI reasoning looks like? You can dive into the technical details and explore the new capabilities at the link below.
Learn more at the official Google DeepMind blog:
No comments