According to CoinMeta, Google has announced that the computer operation tool for Gemini API has concluded its preview phase and is now officially available for use. Compared to the previous preview version, there are no significant changes in functionality or interface; the main improvement is in the stability of the production environment. This tool allows AI to identify screen content through screenshots and then generate operation commands such as clicking, typing, and scrolling. The developer's program is responsible for executing these operations and sending new screenshots back to the model, enabling AI to continue with tasks such as automatically filling out web forms, testing website functions, or collecting information across multiple websites.