Google Antigravity 2.0: Revolutionizing Coding with Gemini

Google recently launched version 2.0 of its agentic coding application, Antigravity, bringing significant improvements with an updated desktop application and a new CLI tool. This update also includes an SDK that allows for workflow customization.

The Antigravity 2.0 desktop application now enables users to manage multiple agents simultaneously and create custom sub-agent workflows. Tasks can be scheduled to run automatically in the background, facilitating integration with Google AI Studio, Android, and Firebase.

The Gemini 3.5 Flash model, developed in collaboration with Antigravity, is at the heart of these new features. Google has also introduced native support for voice commands, a feature already present in other products like Gmail and Docs.

For developers who prefer the terminal, Google has launched a new Antigravity CLI tool, replacing the old Gemini CLI tool. Users are encouraged to migrate to this new interface to create agents more efficiently.

Additionally, Google has introduced an Antigravity SDK, allowing developers to design custom agents. Google Cloud customers can now connect to Antigravity to develop their projects, with custom agent templates available in AI Studio.

Google has also added an export tool to AI Studio, enabling developers to export their projects for local work. Antigravity is used in consumer products like Search, where it provides a real-time customized user interface.

Finally, Google has launched a new AI Ultra plan at $100, offering AI capabilities five times greater than the Pro plan. The price of the AI Ultra plan has been reduced from $250 to $200, allowing for limits twenty times higher than the Pro plan. This pricing strategy aligns with similar offerings from other AI labs like Anthropic and OpenAI.