Gemini

What is Gemini?#

Gemini is Google's multimodal AI assistant, powered by the Gemini 2.5 model family. It supports text, image, audio, and video inputs, integrates with Gmail, Docs, Drive, and Calendar through Workspace, and adds deep reasoning via Gemini Advanced. Extensions connect Gemini to Maps, YouTube, and other Google services.

What can Gemini do?#

• 01

  Deep reasoning across domains

  State-of-the-art reasoning for math, science, and complex analysis.

• 02

  Multimodal understanding

  Works with text, images, video, audio, and code in one model.

• 03

  Agentic coding with tool use

  Follows instructions across multi-step coding tasks and tool calls.

• 04

  Long context window

  Handles extensive codebases and long documents in one session.

• 05

  Deep Research reports

  Runs multi-step web research and compiles a sourced report.

• 06

  Deep Think reasoning mode

  Extended reasoning mode on Ultra for hard technical problems.

• 07

  Video generation with Veo 3.1

  Generate short videos from prompts; tiers vary by plan.

• 08

  Canvas, Gems, and Gmail/Docs integration

  Workspace-wide writing, coding canvas, and custom Gems.

Use Cases#

• “Developers — Debug or refactor a large codebase by pasting it into Gemini 3 Pro and walking through multi-file changes with tool use.
• “Researchers — Run Deep Research on a topic and receive a structured report with citations in minutes.
• “Students — Upload lecture slides, notes, or PDFs and ask Gemini to explain concepts step by step.
• “Creative professionals — Draft a short video from a text prompt with Veo 3.1, then iterate on the scene.
• “Knowledge workers — Use Gems and Canvas inside Gmail and Docs to draft, summarize, and answer with context.