Google Antigravity
AI-assisted coding environment
From Wikipedia, the free encyclopedia
Google Antigravity is an AI-powered integrated development environment (IDE) developed by Google, designed for prioritizing AI agents platform for software development. Announced on November 18, 2025 alongside the release of Gemini 3, Antigravity enables developers to delegate complex coding tasks to autonomous AI agents powered primarily by Google's Gemini 3.1 Pro[5] and Gemini 3 Flash models.[6] The platform is a heavily modified fork of Visual Studio Code. There is debate as to whether it is a direct fork of the software, or whether it is a fork of Windsurf, another AI-oriented code editor which is itself a fork of Visual Studio Code.[7] Google Antigravity supports multiple AI models, including Anthropic's Claude Sonnet 4.6 and Claude Opus 4.6, as well as an open-source variant of OpenAI models (GPT-OSS-120B).[8]
| Google Antigravity | |
|---|---|
| |
 With Agent Side Panel opened | |
| Developer | |
| Initial release | 1.11.2 public preview[1]/November 18, 2025 |
| Preview release | |
| Platform | 64-bit Windows 10 or later, macOS Monterey 12 or later, 64-bit Linux with glibc 2.28 or later and glibcxx 3.4.25 or later[3][4] |
| Type | IDE, artificial intelligence |
| License | Proprietary (free during preview) |
| Website | antigravity |
Antigravity was released in public preview on the day of its announcement and is available free of charge for Microsoft Windows, macOS, and Linux, with "generous rate limits" for Gemini 3.1 Pro usage.[5]
Features
Antigravity introduces an "agent-first" paradigm, shifting from traditional AI code assistance to a system where AI agents operate with greater autonomy. It features two primary views:
- Editor view: A usual IDE interface like Visual Studio Code and PyCharm with an agent sidebar, similar to tools like Cursor or GitHub Copilot.
- Manager view: A control center for orchestrating multiple agents working in parallel across workspaces, allowing asynchronous task execution.[9]
To build user trust, agents generate "Artifacts"—verifiable deliverables such as task lists, implementation plans, screenshots, and browser recordings—rather than raw tool calls.[10] Agents have direct access to the editor, terminal, and a integrated browser, and can learn from previous interactions.[11]