Google has launched Gemini 3.1 Pro, an advanced iteration of its flagship AI model. This upgrade significantly boosts core reasoning and complex problem-solving abilities across consumer, enterprise, and developer applications.
Gemini 3.1 Pro more than doubles the ARC-AGI-2 benchmark score of its predecessor, achieving a verified 77.1%. This score reflects an enhanced capacity to solve novel logic patterns, positioning the model for intricate, multi-step reasoning workflows.
The new model is now available in preview for developers via the Gemini API and through Google AI Studio, Gemini CLI, Google Antigravity, and Android Studio. Enterprise users can access Gemini 3.1 Pro through Vertex AI and Gemini Enterprise. Consumers will find the upgraded model within the Gemini app and NotebookLM for Google AI Pro and Ultra subscribers.
This release follows last week's Gemini 3 Deep Think update, which focused on scientific and engineering applications. Google intends to utilize the preview phase to refine agentic workflows before a wider public release.