Service update · Google
Gemini API's Computer Use tool is now generally available
- Announcement time
- Review updated
It covers both the Interactions API and the GenerateContent API, and works in browser, Android and desktop environments. The model looks at screenshots and suggests clicks and typing, and your code carries them out. Supported models are Gemini 4 Argon, Gemini 3.8 Flash (Google's pick), Gemini 3.5 Flash-Lite and Gemini 3 Flash Preview. There's no separate fee: you pay normal token prices for the model you use, and it isn't available on the free tier. Some actions come back flagged as needing user confirmation or blocked, and your app has to handle that. The old Gemini 2.5 Computer Use preview model has been retired, and Google suggests Gemini 3.8 Flash instead.