Google Launches Gemini 3.1 Pro, Says It Delivers 'Smarter Reasoning' for Complex Tasks

New Delhi: Google has rolled out Gemini 3.1 Pro, the latest model in its AI lineup, describing it as a system designed for situations “where a simple answer isn’t enough.” The release follows a rapid series of updates: Gemini 3 Pro in November 2024, Gemini 3 Flash in December, and Gemini 3 Deep Think last week.

This marks the first time Google has introduced a .1 increment in its Gemini series. Previous mid-cycle updates carried a .5 designation (such as Gemini 2.5 Pro in March 2024). The company says the new approach reflects a focus on refining reasoning capabilities rather than waiting for larger, sweeping upgrades.

 What’s New in Gemini 3.1 Pro

Google claims the model delivers significant gains in reasoning performance. On the ARC-AGI-2 benchmark, which tests a model’s ability to solve unfamiliar logic problems, Gemini 3.1 Pro scored 77.1% which is more than double the performance of Gemini 3 Pro.

Other features include:

- Code-based animation: Ability to generate animated SVGs directly from text prompts, producing scalable, lightweight visuals compared to traditional video.

- Expanded rollout: Available in preview for developers via the Gemini API, AI Studio, Antigravity, Gemini CLI, and Android Studio; for enterprises through Vertex AI and Gemini Enterprise; and for consumers in the Gemini app and NotebookLM.

- Higher usage limits: Pro and Ultra plan subscribers gain expanded access in consumer-facing apps.

 How It Compares to Previous Models

- Gemini 3 Pro: Focused on multimodal capabilities and general performance improvements.

- Gemini 3 Flash: A faster, more efficient variant optimized for lightweight tasks.

- Gemini 3 Deep Think: Tailored for scientific and engineering challenges.

- Gemini 3.1 Pro: Positioned as a smarter baseline model, emphasising reasoning, synthesis, and complex problem-solving across every day and enterprise applications.

Google is releasing Gemini 3.1 Pro in preview mode, with general availability expected after further testing. The company also pointed to ongoing work in agentic workflows, AI systems capable of planning and acting autonomously, suggesting that future Gemini updates may lean more heavily into automation and decision-making.

Read More: 'AI Can Do Many Things, But It Can't Fix Delhi Traffic': Ex-UK PM Sunak