Installation & First Steps
Download, install, and navigate Google Antigravity for the first time
What is Google Antigravity?
Google Antigravity is an AI-powered development environment built by Google DeepMind. It combines a full-featured code editor with an autonomous AI agent that can read, write, and execute code on your behalf.
Think of it as having a senior developer sitting next to you — one who never gets tired, knows every framework, and can simultaneously edit files, run commands, browse the web, and test your application.
Download & Install
Visit antigravity.google/download to get started.
System Requirements
| Platform | Requirements |
|---|---|
| macOS | Monterey (12) or later. Apple Silicon supported. X86 not supported |
| Windows | Windows 10 (64-bit) or later |
| Linux | glibc ≥ 2.28, glibcxx ≥ 3.4.25 (Ubuntu 20+, Debian 10+, Fedora 36+) |
The application automatically notifies you when updates are available — just click “Restart to Update” when prompted.
Basic Navigation: Two Interfaces
Antigravity has two main interfaces that work together:
1. The Editor
The Editor is where you write code. It looks and feels like VS Code — with a file explorer, tabs, terminal, and syntax highlighting. But it’s supercharged with AI features.
Key features:
- Tab Completion — AI-powered inline suggestions as you type
- Command Palette — Quick access to AI actions
- Agent Side Panel — Chat with the agent while you code
- Integrated Terminal — The agent can run commands here
2. The Agent Manager
The Agent Manager is your mission control. It’s where you manage workspaces, view conversations, and orchestrate the agent.
Open it from the Editor with Cmd + E (Mac) or Ctrl + E (Windows).
Key features:
- Workspaces — Organize projects and conversations
- Conversation View — See full agent interaction history
- Panes — Split view for multitasking
- Inbox — Track notifications and updates
- Playground — Experiment without affecting real projects
Switching Between Them
The two interfaces are designed to be used together:
Editor → Agent Manager: Cmd/Ctrl + E
Agent Manager → Editor: Click "Open Editor" or Cmd/Ctrl + EPro Tip: Use the Agent Manager when you want to give the agent a complex, multi-step task. Use the Editor when you want to make quick edits with AI assistance inline.
Your First Interaction
Once installed, open a project folder and try these:
In the Editor
- Open a file and start typing — notice the AI tab completions
- Select some code and press
Cmd/Ctrl + Kto ask the agent to modify it - Open the Agent Side Panel to chat about your code
In the Agent Manager
- Open a workspace with your project
- Type a request like: “Add a dark mode toggle to my app”
- Watch the agent plan, write code, and test it
The agent will:
- Read your existing codebase to understand the project
- Create an implementation plan
- Write the necessary code changes
- Run commands to verify everything works
- Report back with a summary
What’s Next
In the next lesson, we’ll dive deep into the Agent — understanding how it thinks, which models power it, and how to configure its behavior for maximum productivity.