Google previewed a new feature back in May at its I/O 2024 event called Ask about this PDF Gemini overlay which is now widely rolling out on Android. The feature allows users to ask Gemini about a PDF that is opened on the screen. The AI assistant can analyse the PDF and then answer the queries raised by the user.
Spotted by 9to5Google, in the updated PDF viewer within the Files by Google app, activating the Gemini overlay introduces a new “Ask about this PDF” option. Currently, this feature is exclusive to the Files App and isn’t available in Google Drive’s viewer, which continues to display the generic “Ask about this screen” prompt found in other apps. The feature has been spotted in version 16.0.5 of the Google app, which is currently in beta and should soon be released in the stable build of the Google app as well.
When you tap “Ask about this PDF,” the document begins uploading, allowing you to type a query into the “Ask Gemini” field. The send button remains grayed out until the upload is complete. Once processed, the answer is displayed as an overlay with the PDF visible in the background. You can continue asking additional questions afterward.
This process mirrors uploading a file directly to the Gemini app and conducting the conversation there, but the experience is slightly more streamlined here. Document upload is limited to Gemini Advanced customers and leverages the model’s extended context window for processing.
In related news to Gemini, Google recently announced the integration of Gemini in Google TVs for an all-rounded AI experience. The Gemini model on Google TV enables you to do other things like create customised artwork with the family, control your smart home devices while your TV is in ambient mode, and even get an overview of the day’s news. Gemini in Google TVs will begin rolling out later this year on select devices. The company hasn’t specified the exact TV models Gemini would be made available on.