#What are the Key Features of Gemini 3.1 Pro?
Gemini 3.1 Pro represents a significant enhancement from its predecessor, Gemini 3 Pro, chiefly designed to improve reasoning capabilities and problem-solving skills. This upgraded model more than doubles the benchmark score for logic pattern recognition, registering an impressive 77.1% on the ARC-AGI-2 test. Such advancements allow Gemini 3.1 Pro to better handle workflows that demand complex multi-step reasoning, moving beyond simple task execution.
#How Can Developers Access Gemini 3.1 Pro?
The rollout of Gemini 3.1 Pro is currently in preview for developers. They can access this advanced model through the Gemini API within Google AI Studio, as well as through various platforms including Gemini CLI, Google Antigravity, and Android Studio. This access is valuable for developers aiming to incorporate sophisticated AI capabilities into their applications.
#What Should Enterprises and Consumers Expect?
Enterprise users have defined access to Gemini 3.1 Pro via Vertex AI and Gemini Enterprise. On the other side, consumers will encounter this model in the Gemini app and NotebookLM, specifically available for Google AI Pro and Ultra subscribers. This accessibility will empower both segments to leverage enhanced AI functionalities effectively.
#How Does This Upgrade Build on Previous Developments?
This latest version also builds upon the Gemini 3 Deep Think update released recently. The earlier update focused on improving capabilities for scientific, research, and engineering applications. Google intends to use the current preview phase to refine complex agentic workflows, ensuring that these technological improvements can later be made available to a broader audience.