Gemini

Getting Started

1. Visit gemini.google.com and sign in with your Google account — free access to Gemini 2.5 Flash.
2. Upgrade to Gemini Advanced ($20/mo) for 2.5 Pro, the 1M+ token context window, Deep Research, and priority access.
3. For API access, get a key at aistudio.google.com — generous free tier for developers.
4. Try Deep Research: ask Gemini to research a complex topic and it’ll produce a multi-page report with sources.

Model Lineup

• Gemini 2.5 Pro — Google’s most capable model. 1M+ token context, frontier reasoning, strong coding, and the highest scores on many benchmarks.
• Gemini 2.5 Flash — Fast, cost-effective model with a “thinking” mode for enhanced reasoning. Great for most everyday tasks.
• Gemini 2.0 Flash — Previous gen, still available and highly capable for general use.

Key Features

• 1M+ token context window — Analyze entire codebases, long documents, hours of video, or audio in a single prompt. The longest context of any frontier model.
• Native multimodal — Understands text, images, audio, video, and code natively in a single model — not bolted-on capabilities.
• Deep Research — Agentic multi-step research mode that browses the web, synthesizes information, and produces comprehensive reports with citations.
• Google Search grounding — Answers are backed by real-time Google Search results, providing factual, up-to-date information with source links.
• Gems (custom experts) — Create custom AI specialists with persistent instructions for specific workflows, domains, or personas.
• Google Workspace integration — Use Gemini directly in Gmail, Docs, Sheets, Slides, and Meet for AI-powered productivity.
• AI Studio and Vertex AI — Developer platforms for building with Gemini — from quick prototypes in AI Studio to production deployments on Vertex.
• Android and Chrome integration — Gemini is built into Android as the default assistant and available in Chrome for on-page AI assistance.