Select Language:
Google has recently announced updates to its Circle to Search feature. Now, it examines not only the specific area you circle but the entire image to gather more context. This enhancement enables the system to identify multiple objects within a single picture simultaneously, providing a broader set of results.
The latest version of Circle to Search allows users to explore several items within an image at once. The system displays more visual results from a single query, opening new opportunities for merchants and businesses to increase their visibility. For example, when shopping online, rather than searching for a single item like shoes, a user can now look up an entire outfit and get results for all its components.
The functionality relies on the advanced capabilities of Gemini 3’s agentic planning, reasoning, and tool handling. The technology enhances visual query expansion by intelligently analyzing the image to identify its most important parts. It proceeds to perform multiple searches simultaneously across those components and cross-references the findings to generate comprehensive responses. These responses include relevant images seen across the web for each search element.
Demonstrations show users can circle an entire outfit and select a feature to “find the look,” which breaks down the components—such as accessories, shoes, and clothing—into a single, helpful response. Under the hood, Gemini 3 boosts multi-object recognition by splitting the image into parts and running several distinct searches concurrently. This update is available on devices like the Pixel 10 and Samsung Galaxy S26.
ChatGPT
Perplexity
Gemini AI
Grok AI
