Article

Google Antigravity opens browser control: AI can now open webpages, test code, and run multiple tasks in parallel

Beating AI Flash News: Google is gradually rolling out browser control features to Antigravity users. AI can now directly operate the software’s built-in browser to open webpages, click buttons, fill out forms, and view console errors and network requests.

Antigravity previously supported controlling external Chrome via `/browser`. This time, browser operations have been integrated directly into the app, requiring no additional configuration. For example, after writing a website, AI can open the page to test it; if buttons fail or there are layout issues, it can modify the code and recheck.

The new feature also supports multiple browser Agents working in parallel. Each Agent has its own independent tab, allowing users to look up information and test different pages simultaneously. Users can see AI’s mouse actions in real time and take over at any moment.

Google has also added permission controls. The browser uses an independent login environment; when encountering logins or verification codes, it pauses to allow user-initiated actions. When AI initiates file downloads, approval rules still apply, preventing unauthorized downloading of programs or scripts.

Currently, this feature supports macOS, Windows, and Linux, requiring Antigravity App version 2.30.0 or later, and is being gradually rolled out to free, Plus, Pro, and Ultra personal accounts. Enterprise accounts and remote Agents cannot yet use the new built-in browser control.

Original article link https://m.theblockbeats.info/flash/376987