Coinciding with its announcement of Gemini 3 today, Google announced the launch of a new agentic development platform, Google Antigravity.
The company sees Antigravity as an evolution of the IDE into an agent-first future, with capabilities like browser control and asynchronous interaction patterns.
“With models like Gemini 3, we have started hitting the point in agentic intelligence where models are capable of running for longer periods of time without intervention across multiple surfaces. Not yet for days at a time without intervention, but we’re getting closer to a world where we interface with agents at higher abstractions over individual prompts and tool calls. In this world, the product surface that enables communication between the agent and user should look and feel different – and Antigravity is our answer to this,” Google wrote in a blog post.
According to Google, this new IDE is based on four key tenets of collaborative development: trust, autonomy, feedback, and self-improvement.
To facilitate trust, Antigravity produces Artifacts as it works, which are in formats that are more easily verified by developers than raw tool calls, such as task lists, implementation plans, walkthroughs, screenshots, and browser recordings. “Agents in Antigravity use Artifacts to communicate to the user that it understands what it is doing and that it is thoroughly verifying its work,” the company wrote.
For autonomy, Antigravity’s default editor view includes a complete IDE experience, with Tab completions, in-line Commands, and agents in the side panel. There is also an agent-first Manager surface, which can be used to create, orchestrate, and observe multiple agents across multiple workspaces. These two views are intentionally not grouped into a single view, but rather designed as separate spaces with consideration for easily switching between them, Google explained.
The company also believes that one of the failings of agents is the inability to iterate with them. It said that if an agent can complete 80% of a task, but doesn’t provide a way for developers to offer feedback, then they often have to spend more time than the agent saved to finish the last 20%. Antigravity allows developers to provide asynchronous feedback across any surface and Artifact in a number of ways, such as Google Doc-style comments on text Artifacts or select-and-comment feedback on screenshots. The agents then take this feedback and incorporate it back into their execution without needing to be stopped.
Similarly, Google built self-learning into Antigravity, allowing it to retrieve from and contribute to development knowledge bases. It can learn from past work, including explicit information like code snippets or derived architecture and abstract information like the series of steps taken to complete a task.
Google Antigravity is available now in public preview for free, with access to Gemini 3, Anthropic’s Claude Sonnet 4.5, and OpenAI’s GPT-OSS.
