Exploring the Future: How Antigravity Transforms Coding with Gemini 3

Google has unveiled Antigravity alongside the launch of Gemini 3 Pro, a cutting-edge development tool that leverages Gemini 3 Pro and other third-party models. Antigravity is designed to support multiple agents and provide them with direct access to essential tools such as the editor, terminal, and browser, heralding an “agent-first future.”

A key feature of Antigravity is its unique approach to reporting. As tasks are completed, the tool generates Artifacts, including task lists, plans, screenshots, and browser recordings. These Artifacts serve to validate the work done and outline future actions. Google emphasizes that Artifacts offer a more user-friendly verification method compared to traditional lists of actions and tool calls.

Antigravity introduces two primary usage views: the default Editor view, providing a familiar Integrated Development Environment (IDE) experience with an agent in a side panel, and the new Manager view, designed for overseeing multiple agents concurrently. This Manager view acts as a centralized hub for managing and monitoring agents across various workspaces simultaneously, akin to “mission control.”

Google has expanded feedback options for AI agents within Antigravity, enabling users to leave comments on specific Artifacts for agents to consider without interrupting their workflow. Additionally, agents can now “learn from past work,” retaining valuable code snippets and task execution steps for future reference.

Antigravity is currently available in a public preview and is compatible with Windows, macOS, and Linux systems. The tool is free to use, offering “generous rate limits” for Gemini 3 Pro, Claude Sonnet 4.5, and OpenAI’s GPT-OSS. Google assures users that rate limits refresh every five hours, with only a select few power users ever reaching these limits.