Google is expanding the capabilities of AI Mode, its experimental feature within Google Search, by introducing powerful image-based search functionality. Users can now ask complex, multi-step questions about photos they’ve taken or uploaded directly into AI Mode.
This new feature is powered by Google Lens’ multimodal technology, which enables AI Mode to analyze entire scenes within an image. It can interpret relationships between objects, their materials, colors, shapes, and layout—offering a deeper understanding than standard image recognition.
Using a process called “query fan-out,” AI Mode generates multiple queries based on both the image and its contents to deliver rich, detailed answers. For instance, snapping a picture of your bookshelf and asking, “If I liked these books, what are some highly rated similar ones?” will prompt AI Mode to identify each title and suggest personalized book recommendations, complete with links for more information or purchase.
The tool also supports follow-up questions, allowing users to refine their searches. For example, you might ask, “Which of these is the shortest read?” to get a quick-reading suggestion from the list of recommendations.
Alongside this update, Google announced it’s expanding access to AI Mode to millions more users through Labs, its hub for experimental features. Previously, AI Mode was limited to Google One AI Premium subscribers.
