Introduction
Circle to Search has been extended to handle several objects in a single picture, letting users identify and ask about each element at once. The feature builds on the original capability that let users tap or scribble to retrieve information, now upgraded for broader queries.
Multi‑Object Detection Workflow
Users simply draw circles around every item they wish to learn about-whether fish in a photo or pieces of an outfit-and submit a single query. The system parses each region, runs parallel searches, and returns a consolidated list of results. This approach provides speed and reduces the need for repeated interactions.
Device Integration
The updated experience launches today on the Samsung Galaxy S26 series and Pixel 10 devices, with additional Android hardware slated to receive the capability. Integration leverages native camera and UI layers, ensuring the circle‑drawing interface feels natural on each device.
Gemini 3 Engine
Behind the scenes, Gemini 3 powers the identification of image parts, executes the searches, and assembles the answers. Its model interprets the visual context of each circled region, enabling precise matching against the indexed web content. The engine delivers accuracy that aligns with user expectations for visual queries.
Practical User Scenarios
Fashion enthusiasts can outline an entire ensemble, retrieve each clothing items details, and even launch a virtual try‑on if the feature is supported in their country. Outdoor photographers may circle wildlife to discover species information, while shoppers can highlight shoes to find similar products. These examples illustrate how the tool adapts to varied interests.
Impact and Adoption
Since its debut, Circle to Search has processed billions of queries per month, supporting common questions such as What are those shoes? or Where is this hiking trail? The multi‑object enhancement promises to increase engagement by allowing richer, compound inquiries, thereby improving the overall utility of the service.