Google Antigravity

From Open Source Ecology
Jump to navigation Jump to search

Template:Infobox software Google Antigravity is an AI-powered integrated development environment (IDE) developed by Google. Released on November 18, 2025, alongside Gemini 3, it is designed to shift software development from a copilot model to an agent-first paradigm, where AI agents operate with high autonomy to plan, execute, and verify tasks.[1][2]

Features

Google Antigravity introduces several key features aimed at enhancing developer productivity through AI autonomy:

  • Agent-First Paradigm: Unlike traditional coding assistants that suggest code snippets, Antigravity employs autonomous agents capable of handling complex engineering tasks such as refactoring, debugging, and feature implementation with minimal human oversight.[3]
  • Dual Views: The IDE provides two primary interfaces:
    • Editor View: A code-centric interface similar to Visual Studio Code, enhanced with an agent sidebar.
    • Manager View: A control center for orchestrating multiple agents working asynchronously across different workspaces.
  • Cross-Surface Control: Agents have synchronized access to the editor, terminal, and browser, allowing them to perform end-to-end tasks like running commands and verifying web application behavior.
  • Artifacts: To maintain transparency, agents generate "Artifacts"—including implementation plans, task lists, and logs—that developers can review and audit.

Development

Google Antigravity was launched in public preview on November 18, 2025. It is powered primarily by Google's Gemini 3 Pro model but supports model optionality, allowing integration with other models such as Claude Sonnet 4.5 and open-source variants.

References