What is it

Gemini is Google's consumer-facing AI assistant powered by the Gemini model family (3.5 Flash and Gemini 3 / 3.1 Pro at the time of writing, with Deep Think available on Ultra). It is natively multimodal — text, image, audio, video, and code — and ships across the gemini.google.com app, Android/iOS apps, AI Mode in Google Search, and Gemini in Chrome, Gmail, Docs, Vids, and other Workspace surfaces.

What it can do

Conduct Deep Research that browses hundreds of websites (and optionally your Gmail / Drive / Chat) into cited multi-page reports, generate images via Nano Banana Pro and videos via Veo 3.1 in Flow, write and edit interactively in Canvas, build reusable "Gems" expert chatbots, summarize and turn sources into Audio Overviews and notebooks in NotebookLM, and reason over long inputs with a 1M token context window on Pro/Ultra.

Who is it for

Knowledge workers and students living in Google Workspace, researchers and analysts who need long-source synthesis, marketing and creative teams using Veo/Imagen/Nano Banana, and developers using the connected Jules async coding agent, Google Antigravity, Gemini Code Assist, and Gemini CLI.

Key Features

Gemini 3 Pro Multimodal Reasoning – text, image, audio, video, code

Gemini 3 / 3.1 Pro is Google's most capable reasoning model, accepting text, images, audio, video, PDFs, and entire code repositories in a single prompt. It powers the Gemini app on web, Android, and iOS, with Ultra subscribers getting first access to Deep Think — the higher-reasoning variant tuned for the hardest problems.

Deep Research Agent – hundreds of sources, multi-page report

Deep Research turns Gemini into an agent that decomposes a query, browses up to hundreds of websites (and optionally your Gmail, Drive, and Chat), iteratively explores fresh links, and synthesizes findings into a cited multi-page report. On Pro and Ultra it runs on the Gemini 3 model with higher monthly limits, and Ultra unlocks Deep Search in AI Mode for the most thorough Google-Search-native variant.

AI Mode and Deep Search in Google Search

Google Search's AI Mode brings Gemini directly into the search box, with Deep Search browsing hundreds of sites and producing a comprehensive, fully cited report in minutes. Higher access to Gemini 3 Pro and agentic capabilities in AI Mode is gated to Pro, with the highest access on Ultra.

Canvas – write, code, and prototype in one workspace

Canvas opens a split-panel workspace inside Gemini for collaborative writing, real-time code editing, and live preview of small web apps, games, and infographics built straight from a prompt or Deep Research report. Canvas is free for all users, but Pro and Ultra unlock the most capable Gemini 3 model and a 1M token context window for larger projects.

Use Cases

Long-form competitive and market research

Use Deep Research to generate competitor deep-dives, industry overviews, and policy briefings from hundreds of sources in minutes, then push the report into Canvas to turn it into a web page, infographic, quiz, or Audio Overview.

Studying and exam prep

Upload notes, slides, and textbooks into NotebookLM or the Gemini app, generate study guides, practice tests, Audio Overviews, and flashcards, and use Canvas to step through problems with the Gemini 3 model.

Workspace productivity

Draft and rewrite emails in Gmail, summarize long Docs, generate slide outlines in Vids, and pull insights out of Sheets, all with Gemini directly in the apps and with Pro/Ultra context across the user's Drive.

Long-document and codebase reasoning

Take advantage of the 1M token context window to load entire books, hour-long video transcripts, or full code repositories (Google quotes up to 30k lines of code) into a single conversation for analysis, refactoring suggestions, and Q&A.

Pricing plans

Frequently Asked Questions

Discussion

No comments yet. Be the first to start the thread.

Gemini