Google Antigravity Freemium

Google Antigravity is a free agent-first IDE, launched in November 2025. Built on VS Code with Gemini 3 as the default model, it lets AI agents take a written task, produce a plan, then edit files, run terminal commands and drive a browser to verify the result.

What is Google Antigravity?

Antigravity is Google’s answer to Cursor and Windsurf, but it pushes further in one direction: the agent, not the editor, is the primary interface. Instead of a chat sidebar, you get an Agent Manager that tracks several agents working on different tasks. Each agent produces artifacts — plans, task lists, screenshots, browser recordings — that you review.

It ships for macOS, Windows and Linux. Underneath it’s a VS Code fork, so existing extensions, keybindings and themes carry over.

Google Antigravity Features

Agent Manager — run multiple agents in parallel across workspaces and inspect each one’s status.

Browser control — agents open a Chromium instance, click through your app and record what they see, so “verify it works” is a real step.

Artifacts — plans, walkthroughs and screenshots stored as reviewable checkpoints rather than chat scrollback.

Model choice — Gemini 3 Pro, Gemini 3 Flash, plus Anthropic’s Claude Sonnet and OpenAI models where available.

Tab completion plus inline edits — the usual editor-side assist, inherited from the VS Code lineage.

Free tier — generous rate limits during the public preview period.

Is Google Antigravity Free?

Yes, at the time of writing. Antigravity launched in public preview at no cost, with usage rate-limited per model rather than priced per token. Google hasn’t published a commercial pricing page yet, so anyone building a business on it should expect that to change. Compare that to Cursor’s $20/month Pro plan or Windsurf’s $15 entry tier, and the free preview is the main reason people are trying it.

What Is Antigravity Good At?

Greenfield features — “build a login flow with tests” as one prompt, with the agent verifying in-browser.

UI bug hunting — the browser step catches rendering issues a terminal-only agent never sees.

Parallel chores — dependency bumps and test fixes running in separate agents at once.

Prototyping — free access to Gemini 3 Pro makes throwaway experiments cheap.

Teams on Google Cloud — natural fit if your deploy target is already GCP.

Who Should Use Antigravity and Who Should Not

Try it if you want agent autonomy without a subscription, or if your work involves verifying browser behaviour. Skip it if you need stability for a production team — this is preview software, and preview software breaks. Skip it too if you’re locked into JetBrains IDEs or need a support contract.

The honest complaint: agent-first workflows are impressive in demos and slower in practice. Reviewing an agent’s twenty-step plan takes real attention, and a wrong assumption near step three wastes everything after it. Keep tasks small.

Quick Summary

Google Antigravity is a free, agent-first IDE launched in public preview in November 2025.

It’s built on VS Code, so extensions and settings carry over.

Agents can edit files, run terminal commands and control a Chromium browser.

Default model is Gemini 3, with Claude and OpenAI options available.

No published pricing yet — free during preview, expect a paid tier later.