Google has begun rolling out Gemini 3.1 Pro, the latest version of its flagship AI model, positioning it as an upgrade focused on improved reasoning and complex problem-solving. The new model is available starting today through the Gemini app for subscribers on AI Pro and AI Ultra plans, as well as in NotebookLM and a range of developer tools.
The release of Gemini 3.1 Pro comes amid an increasingly competitive AI landscape. Model updates are arriving at a rapid pace across the industry, with companies iterating quickly on reasoning performance, multimodal capabilities, and automation features. In this context, Google’s emphasis with Gemini 3.1 Pro is measurable improvement in benchmark results and broader availability across consumer and enterprise products.
One of the headline figures shared by Google is the model’s performance on the ARC-AGI-2 benchmark, a test designed to evaluate abstract reasoning. Gemini 3.1 Pro reportedly achieved a score of 77.1%, a substantial increase over the 31.1% score recorded by Gemini 3 Pro on the same test. While benchmark results don’t always translate directly into real-world outcomes, they offer a standardized way to compare progress between model generations.
Beyond raw scores, Google has highlighted several practical applications. Gemini 3.1 Pro can generate animated SVG files from text prompts, producing scalable graphics suitable for web use without loss of quality. The company also demonstrated creative coding tasks, including generating a full website concept inspired by a literary character, as well as building interactive designs such as a 3D starling murmuration simulation with dynamic sound elements.
These examples reflect a broader push toward multimodal and design-oriented AI workflows. Rather than focusing solely on text generation, Gemini 3.1 Pro is positioned as a tool for developers, designers, and enterprises looking to streamline prototyping and creative production. The model is accessible through the Gemini API, including integrations with AI Studio, Gemini Enterprise, Antigravity, and Android Studio.
For consumers, the upgrade is tied to subscription tiers. Users on AI Pro and AI Ultra plans will see Gemini 3.1 Pro inside the Gemini app and NotebookLM. This approach reinforces Google’s strategy of gating its most advanced AI capabilities behind paid offerings, while still maintaining a presence across its broader ecosystem.
As the AI model landscape continues to evolve, the key question is less about isolated benchmark wins and more about sustained improvements in reliability, reasoning depth, and real-world usability. Gemini 3.1 Pro represents another incremental step in that direction, with Google aiming to strengthen its position in both consumer-facing AI tools and enterprise development environments.
