Google has launched Antigravity, a development tool that leverages its Gemini 3 Pro and third-party models. Currently in free public preview, Antigravity allows users to control their browser and provides documentation of its work through features called Artifacts. These Artifacts include task lists, plans, screenshots, and recordings that verify the completed tasks and outline future steps. This approach is intended to simplify the verification process for users, making it easier to track what the tool is doing. Antigravity offers two main views for users. The default Editor view resembles traditional Integrated Development Environments (IDEs) like Cursor and GitHub Copilot, featuring an agent in a side panel. The Manager view is designed for users managing multiple agents simultaneously, akin to a mission control setup. This view allows for more autonomous operation of agents across different tasks. Users can also provide feedback directly on specific Artifacts, enabling agents to adjust their work based on past performance without interrupting their current tasks. The tool is compatible with Windows, macOS, and Linux, and Google claims it has "generous rate limits" for using Gemini 3 Pro, refreshing every five hours. While it can handle various AI models, including Claude Sonnet 4.5 and OpenAI’s GPT-OSS, Google notes that only a small number of power users might hit those limits, making it broadly accessible for everyday developers.