What is the ARC-AGI-2 benchmark score for Gemini 3.1 Pro?

Gemini 3.1 Pro achieved a verified score of 77.1% on the ARC-AGI-2 benchmark, which evaluates a model's ability to solve entirely new logic patterns. This represents more than double the reasoning performance of Gemini 3 Pro.

How can developers access Gemini 3.1 Pro?

Developers can access Gemini 3.1 Pro in preview through the Gemini API in Google AI Studio, Gemini CLI, the agentic development platform Google Antigravity, and Android Studio. Enterprise users can access it through Vertex AI and Gemini Enterprise.

What are the key improvements in Gemini 3.1 Pro compared to 3 Pro?

Gemini 3.1 Pro represents an upgrade in core reasoning capabilities over 3 Pro. While 3 Pro focused on general capabilities, 3.1 Pro specifically enhances complex problem solving, achieving significantly higher scores on rigorous reasoning benchmarks like ARC-AGI-2 and improved performance on agentic workflows.

Is Gemini 3.1 Pro available for free?

In the Gemini consumer app, 3.1 Pro is available with higher rate limits for users with Google AI Pro and Ultra plans. A limited version may be accessible to free users with standard rate limits. For developers, it is currently available in preview with specific pricing details to be announced upon general availability.

What types of tasks is Gemini 3.1 Pro best suited for?

3.1 Pro is designed for tasks where simple answers aren't enough, including complex data synthesis, generating code based animations and interactive designs, explaining complex topics visually, and supporting agentic workflows that require advanced reasoning and tool use.

Gemini 3.1 Pro

Name: Gemini 3.1 Pro
Brand: Google
Rating: 4.5 (128 reviews)

Google's Gemini 3.1 Pro delivers advanced reasoning with 77.1% ARC-AGI-2 score. Access via API, Vertex AI, and Gemini app for complex problem-solving and creative coding.

Visit Gemini 3.1 Pro View Media

About Gemini 3.1 Pro

Gemini 3.1 Pro represents Google's latest advancement in foundational AI model capabilities, building upon the Gemini 3 series released in November. This upgraded core intelligence delivers significantly enhanced reasoning performance, achieving a verified 77.1% score on the ARC-AGI-2 benchmark. It is more than double the performance of its predecessor, Gemini 3 Pro.

The model is specifically architected for complex tasks where simple pattern recognition proves insufficient. Unlike standard conversational AI, 3.1 Pro excels at synthesizing disparate data sources into coherent insights, generating production ready code assets like animated SVGs, and constructing interactive simulations requiring multi step logical deduction. Google has positioned this release as a "smarter baseline" for agentic workflows, emphasizing its utility in research, engineering, and creative prototyping.

Currently available in preview across Google's ecosystem, including the Gemini API, Vertex AI, and consumer applications like NotebookLM—3.1 Pro signals Google's push toward more capable, reasoning focused AI systems. The model supports advanced use cases ranging from aerospace telemetry visualization to literary themed creative coding projects.

Key Features

Advanced Reasoning: Achieves 77.1% on ARC-AGI-2 benchmark for novel logic pattern solving, representing a 2x improvement over Gemini 3 Pro
Code Based Animation: Generates website ready animated SVGs directly from text prompts with pure code output for scalability and small file sizes
Complex System Synthesis: Bridges complex APIs and user friendly design by configuring data streams into functional dashboards like aerospace telemetry visualizations
Interactive Design: Creates immersive 3D experiences with hand tracking capabilities and generative audio that responds to visual elements in real time
Creative Coding: Translates literary themes and atmospheric tones into functional web interfaces and contemporary digital experiences
Multi Platform Access: Available across Gemini API, Vertex AI, Google AI Studio, Gemini CLI, Antigravity, Android Studio, and consumer apps

Pricing

Consumer Access: Included with Google AI Pro and Ultra plans Higher rate limits available exclusively for Pro and Ultra subscribers in Gemini app and NotebookLM.
Developer/Enterprise Preview: Available via API and cloud platforms Currently in preview through Gemini API in Google AI Studio, Vertex AI, Gemini Enterprise, Gemini CLI, Antigravity, and Android Studio. General availability pricing to be announced.

Pricing last updated: February 23, 2026 at 6:16 AM

Use Cases

Complex problem solving requiring multi step reasoning and synthesis
Data visualization and interactive dashboard creation from API sources
Creative coding projects including generative animations and 3D experiences
Research and engineering workflow automation with agentic capabilities

Pros & Cons

Pros:

State of the art reasoning capabilities with 77.1% ARC-AGI-2 verification
Extensive platform availability across Google's developer and consumer ecosystem
Advanced code generation producing executable visual and interactive assets

Cons: