01
Google makes Computer Use generally available in the Gemini API
On 30 October Google moved Computer Use, the Gemini API tool that lets an agent operate a browser, an Android phone or a desktop by reading screenshots and returning clicks and keystrokes, from preview to general availability in both its Interactions and GenerateContent APIs. Google's documentation lists four supported models and recommends Gemini 3.8 Flash for the task. According to the pricing page, the tool is not available on the free tier; on the paid tier it is billed at the normal token rates of the chosen model.
What it means for you
General availability makes it more realistic to build agents that fill in forms or move data between web tools that have no API. If you test it, follow Google's own advice: run the agent in an isolated browser or virtual machine, and require human confirmation before purchases, sending messages or accepting terms.