Qu'est-ce que Google Antigravity ?
Google Antigravity launched in November 2025 as an AI-powered code editor, then relaunched as Antigravity 2.0 at Google I/O in May 2026 as a full standalone agent-first development platform: a desktop app for orchestrating multiple coding agents at once, a CLI (which absorbed and replaced the earlier Gemini CLI), an SDK, a Managed Agents API, and an enterprise deployment path.
The desktop app lets developers run multiple agents in parallel, design custom subagent workflows, and schedule tasks to run automatically in the background rather than working through one task at a time in a single chat window. It's powered by Gemini 3.5 Flash, which Google says outperforms its own Gemini 3.1 Pro on most benchmarks while running roughly four times faster, and includes live voice transcription via Gemini Audio models.
Antigravity includes a free tier for individual use. Higher usage requires a Google AI Pro or the new AI Ultra plan, which costs $100/month and provides 5x the usage limits of the standard Pro tier.
Fonctionnalités clés de Google Antigravity
- Multi-agent orchestration: Run and coordinate multiple coding agents simultaneously from one desktop app.
- Antigravity CLI: A server-side, terminal-based agent harness that replaced the earlier Gemini CLI.
- Custom subagent workflows: Design and schedule background-running task workflows.
- SDK and Managed Agents API: Build agent-first workflows into your own tools and infrastructure.
- Live voice transcription: Voice input support powered by Gemini Audio models.
Cas d'usage de Google Antigravity
- Run multiple coding agents in parallel on different parts of a project
- Automate a background-running development workflow on a schedule
- Build a custom agent-first tool on top of the Antigravity SDK or Managed Agents API
- Use the Antigravity CLI for terminal-based agentic development