Google has rolled out significant updates for its AI assistant Gemini, introducing a new tool in Chrome and enhancing the Gemini 3.5 Flash model with computer use capabilities. These improvements represent a leap forward in integrating artificial intelligence into web browsing and software development.
Chrome 149 Brings 'Select from Screen' for Quicker Gemini Interactions
The latest version of the Chrome browser, version 149, introduces a 'Select from screen' tool for Gemini users. Located in the plus menu at the bottom, this tool allows users to highlight any text or image from the active tab and send it directly as context for a Gemini prompt. Simply click on the 'Select from screen' option, choose the area of interest, and the content is automatically added to the prompt box. The feature is rolling out gradually; if not yet visible, a browser restart should activate it. This tool simplifies using Gemini for webpage analysis, information extraction, or generating summaries based on visual content.
Sponsored Protocol
Gemini 3.5 Flash Natively Integrates Computer Use
In parallel, Google announced that the Gemini 3.5 Flash model now includes a built-in computer use tool, replacing the previous standalone Gemini 2.5 Computer Use model. This native integration joins existing grounding for Search and Maps. With this new capability, developers can create custom AI agents that can see, reason, and act across browser, desktop, and mobile environments. Performance has been optimized for long-horizon tasks and enterprise automation, enabling use cases like continuous software testing and knowledge work on professional applications.
Google provided a concrete example: the 3.5 Flash model uses computer use to analyze the Gemini app itself and return a categorized list of features. This demonstrates the model's ability to interact with complex user interfaces. On the safety front, the company assured that enterprise customers can rely on robust protections when using this technology.
Sponsored Protocol
Gemini 3.5 Flash with computer use is available immediately via the Gemini API. Developers interested can start integrating these capabilities to build advanced automations. For those wanting to explore the implications of data usage by Google, we recommend reading our article on Google Enables Automatic Media Saving for AI Training, which covers related privacy and AI training topics.
These updates confirm Google's strategy to make AI more pervasive and useful. While the 'Select from screen' tool enhances the user experience in Chrome, Gemini 3.5 Flash's computer use opens the door to new forms of automation. For a comprehensive overview of Google's language models, see the Wikipedia page on Gemini.
Source: https://9to5google.com/2026/06/24/gemini-chrome-select-screen
