8 Essential Ways to Use Gemini 3.1 Pro Within 24 Hours

Google released Gemini 3.1 Pro, claiming double the inference performance of its predecessor. In the ARC‑AGI‑2 benchmark it achieved a 77.1% score, roughly twice the previous Gemini 3 Pro, and in the MCP‑Atlas complex‑task benchmark it reached 69.2%, surpassing Claude Opus 4.6 by nearly ten points.

1. Quick Experience: Web UI + Mobile App

Open https://gemini.google.com/ in a browser or the mobile app. Registration with a Google account provides immediate access. The web version now runs Gemini 3.1 Pro, while the app rollout is ongoing. Free users receive about five queries per day; Pro ($19.99 / month) and Ultra ($249.99 / month) subscriptions grant higher limits.

2. NotebookLM – AI‑Powered Note‑Taking

Visit https://notebooklm.google/. The tool ingests dozens of papers, PDFs, or web links, then extracts core ideas, builds logical connections, and creates a structured knowledge base. It offers two main features:

Audio Overview : AI hosts generate a conversational podcast from the uploaded material.

Video Overview : AI‑narrated dynamic slides with auto‑generated graphics via Nano Banana.

Gemini 3.1 Pro is available on NotebookLM for Google AI Pro and Ultra subscribers.

3. Google Search AI Mode – Search‑Integrated Reasoning

On the Google homepage, toggle the “AI Mode” next to the search box. Queries are fed directly to Gemini 3.1 Pro, which returns structured, in‑depth analyses that combine real‑time web results with the model’s reasoning capabilities. This is ideal for complex, up‑to‑date questions such as “What AI open‑source projects will matter in 2026?”

4. Google AI Studio – No‑Code App Builder

Access https://aistudio.google.com/. The platform lets users describe desired functionality in natural language; Gemini 3.1 Pro then generates a full application prototype. The standout “Annotation Mode” lets users draw circles and add spoken instructions to modify the app precisely. Free tier users enjoy a generous quota and a full‑size 1 M‑token context window.

Warning: The service collects usage data; avoid sensitive projects.

5. AI Studio API – Personal Developer Access

Generate an API key at https://aistudio.google.com/api-keys. Pricing is token‑based: for standard context (≤200 k tokens) input costs $2 and output $12; for extended context (>200 k tokens) input $4 and output $18.

6. Gemini CLI – Command‑Line Interface

Install the CLI with npm install -g @google/gemini-cli and run gemini to invoke the model directly from the terminal.

7. Antigravity – Google’s AI‑Powered IDE

Visit https://antigravity.google/. Antigravity offers a “Google‑version of Cursor,” supporting Gemini 3.1 Pro, Claude Sonnet 4.6, and others as interchangeable agents. It excels at multi‑step workflows, code generation, debugging, refactoring, and documentation, outperforming simple code‑completion tools.

8. Vertex AI – Enterprise‑Grade Cloud Platform

Explore https://cloud.google.com/vertex-ai. Vertex AI provides a full MLOps stack—training, tuning, deployment, monitoring, and A/B testing—with options for cloud or private deployments, data isolation, and compliance. New users receive $300 credit for three months; thereafter usage is pay‑as‑you‑go.

Overall, Gemini 3.1 Pro shines in multi‑step, complex workflows, offering superior benchmark scores, extensive multimodal capabilities, and a growing ecosystem of tools for individuals, researchers, and enterprises.