Google Launches Gemini 3.1 Pro With Advanced Reasoning Capabilities
Google has officially introduced Gemini 3.1 Pro, an upgraded artificial intelligence model built to handle complex problem-solving across science, research, and engineering.
The new version expands on the Gemini 3 series and delivers major improvements in core reasoning performance.
Gemini 3.1 Pro Doubles Reasoning Performance
On the ARC-AGI-2 benchmark — a test that measures a model’s ability to solve unfamiliar logic patterns — Gemini 3.1 Pro achieved a verified score of 77.1%. This result more than doubles the reasoning performance of the previous Gemini 3 Pro model.
The benchmark highlights the model’s enhanced ability to analyze novel problems, recognize patterns, and deliver accurate solutions in advanced use cases.
Developer and Enterprise Access Across Google Platforms
Starting Thursday, Gemini 3.1 Pro is rolling out in preview to developers through multiple Google platforms. These include:
-
The Gemini API in Google AI Studio
-
Gemini CLI
-
Google Antigravity, the company’s agentic development platform
-
Android Studio
Enterprise customers can access the model via Vertex AI and Gemini Enterprise. Meanwhile, consumers can use Gemini 3.1 Pro through the Gemini app and NotebookLM.
Users subscribed to Google AI Pro and Ultra plans benefit from higher usage limits in the Gemini app. NotebookLM access to Gemini 3.1 Pro is reserved exclusively for Pro and Ultra subscribers.
Built for Complex System Synthesis and Advanced Tasks
Gemini 3.1 Pro is optimized for tasks that require deep reasoning and structured thinking. These include generating visual explanations of technical subjects, synthesizing large datasets, and supporting advanced creative workflows.
One standout capability is “complex system synthesis.” This allows the model to connect sophisticated APIs with intuitive design interfaces. Google demonstrated this feature by building a live aerospace dashboard that tracks and visualizes the International Space Station’s orbit in real time.
Preview Release Ahead of Wider Rollout
Google described the current launch as a preview phase aimed at validating the model’s performance before broader availability. Additional enhancements are expected, particularly in agentic workflows and advanced automation capabilities.
With Gemini 3.1 Pro, Google continues to strengthen its position in the rapidly evolving AI landscape, focusing on deeper reasoning, developer tools, and enterprise-grade applications.
