Google released Gemini 3 Pro on November 19, and today, exactly three months later, it introduces Gemini 3.1 Pro. According to Google, this model "represents a step forward in core reasoning," making it "a smarter, more capable baseline for complex problem-solving."
This advancement is evident in its performance across various AI benchmarks. The image below compares Gemini 3.1 Pro with other recent models from Google's competitors and with Gemini 3 Pro.
Google emphasizes Gemini 3.1 Pro's impressive performance in the ARC-AGI-2 benchmark, which evaluates a model's capacity to address entirely new logic patterns. Gemini 3.1 Pro excelled in this evaluation.
The new model is specifically designed for scenarios where simple answers fall short, employing "advanced reasoning" to tackle your most challenging tasks. Gemini 3.1 Pro's enhanced intelligence facilitates practical applications, whether you need a clear visual explanation of a complex topic, a method to consolidate data into a single view, or assistance in bringing a creative project to fruition.
Starting today, Gemini 3.1 Pro is being rolled out through the Gemini app, with Google AI Pro and Ultra subscribers receiving higher limits. Additionally, Gemini 3.1 Pro will be available in NotebookLM, though only to subscribers of Google AI Pro and Ultra plans.
Source
