Google Antigravity 2.0: From Code Editor to Autonomous Agent Orchestrator

Breaking News: Google Reinvents Antigravity as Agent Orchestration Platform

Six months after its initial launch as a coding environment, Google today announced a sweeping expansion of Antigravity — repositioning it as a full-fledged platform for developing and managing teams of autonomous AI agents. The move, unveiled ahead of Google I/O 2026, introduces a standalone desktop app, a command-line interface (CLI), a software development kit (SDK), and deep enterprise cloud integration.

“We are expanding Antigravity beyond a coding environment and turning it into a platform to develop and manage teams of autonomous AI agents,” said Koray Kavukcuoglu, CTO of Google DeepMind and Chief AI Architect at Google, during a press briefing this week. The company claims millions of developers are already building with the platform.

The announcement marks a strategic pivot from a single-user coding tool to a multi-agent coordination hub, positioning Antigravity as a competitor to emerging agent platforms from OpenAI, Microsoft, and others.

Platform Details: Desktop, CLI, SDK, and Cloud

The centerpiece is Antigravity 2.0, a standalone desktop application that serves as a central hub for agent orchestration. Developers can now coordinate multiple agents working simultaneously — one generating code, another producing brand assets, a third handling product architecture planning. “Google Antigravity is our agent-first development platform for developers to take an idea and turn it into a production-ready app,” write Varun Mohan, Director of Software Engineering at Google DeepMind, and Logan Kilpatrick, Member of the Technical Staff, in a blog post.

For terminal enthusiasts, the Antigravity CLI offers a lightweight interface for spinning up agents without a graphical environment. Existing Gemini CLI users are encouraged to migrate. Rounding out the expansion, the Antigravity SDK provides programmatic access to the same agent harness powering Google’s internal products.

The SDK is co-optimized for Gemini models, specifically Gemini 3.5 Flash — described as the platform’s primary workhorse model. Google states it outperforms Gemini 3.1 Pro on almost all benchmarks and runs four times faster than competing frontier models.

Cloud & Enterprise Integration

Managed Agents in the Gemini API extend the platform into the cloud. A single API call can now spin up a full agent that reasons, uses tools, and executes code inside a persistent, isolated Linux environment. State and files persist across follow-up calls, enabling multi-turn sessions without re-initialization. “Today, we’re expanding the Antigravity ecosystem to manage and deploy agents that can integrate across key developer surfaces,” Mohan and Kilpatrick write.

The enterprise angle is addressed through Antigravity’s integration with the Gemini Enterprise Agent Platform, enabling Google Cloud customers to connect directly to their Cloud projects. A new $100-per-month Google AI Ultra subscription targets heavy users.

Background

Launched at Google I/O 2025, Antigravity initially served as a browser-based coding environment for rapid prototyping with AI assistance. Its original focus was on code generation and debugging within a single workspace. The platform gained traction among early adopters but remained niche compared to established IDEs.

Over the past six months, Google has quietly expanded Antigravity’s capabilities, integrating it deeper with its Gemini model family and adding multi-turn agent support. The shift to agent orchestration — where multiple autonomous agents collaborate on complex workflows — reflects a broader industry trend toward “agentic” AI architectures. Google’s investment in Antigravity signals that it sees agent management as a key battleground for developer mindshare.

What This Means

For developers, this transformation turns Antigravity from a coding convenience into a production-ready platform for building and deploying autonomous AI systems. The ability to manage teams of agents — each specialized for different tasks — opens new possibilities for automating complex workflows, from software development to creative production.

Enterprise teams gain a unified way to deploy agents within their existing Google Cloud infrastructure, while individual developers can experiment with multi-agent coordination through the free CLI and SDK. However, the added complexity may require a learning curve, and the $100/month Ultra subscription could be a barrier for smaller teams.

Long-term, Google’s bet on Antigravity as an agent-first platform positions it to compete directly with emerging orchestration tools from rivals — potentially making it a critical piece of the AI development stack. For Google, the move also deepens lock-in to its Gemini ecosystem, encouraging developers to build agents that rely on Google’s models and cloud services.