Google released Gemini 3 Pro on November 19, and today, exactly three months later, it’s introducing Gemini 3.1 Pro. This model “represents a step forward in core reasoning”, Google says, being “a smarter, more capable baseline for complex problem-solving”.
This, Google says, is reflected in its performance in various AI benchmarks. You can see Gemini 3.1 Pro compared to the other latest models from Google’s competitors, as well as with Gemini 3 Pro, in the image below.
Google highlights Gemini 3.1 Pro’s performance in ARC-AGI-2, which it describes as “a benchmark that evaluates a model’s ability to solve entirely new logic patterns”. As you can see, Gemini 3.1 Pro came out on top of that one.
Google says this new model is “designed for tasks where a simple answer isn’t enough”, since it uses “advanced reasoning” for “your hardest challenges”. Gemini 3.1 Pro’s improved intelligence can help in practical applications, when “you’re looking for a clear, visual explanation of a complex topic, a way to synthesize data into a single view, or bringing a creative project to life”.
Gemini 3.1 Pro is rolling out to everyone starting today via the Gemini app, but Google AI Pro and Ultra subscribers will get higher limits. Gemini 3.1 Pro is also available in NotebookLM, but only for subscribers to Google AI Pro and Ultra plans.
