Pinterest expands AI visual search tools to refine image-based discovery and shopping

Pinterest has introduced a set of updates to its visual search capabilities, aiming to make it easier for users to search, refine, and shop for fashion content directly from images. The new features are rolling out in the women’s fashion category across the U.S., Canada, and the U.K., with plans to expand into additional categories and regions.

The latest tools use AI models, including Visual Language Models (VLMs), to generate descriptive keywords based on elements in a Pin. When users tap on an image, an animated glow highlights areas of interest, and keywords are automatically generated to help users identify and search for similar items. This function is intended to assist users who may not have the words to describe the visual elements they’re interested in, such as particular styles or aesthetics.

Pinterest has also added a refinement bar that allows users to narrow down search results based on attributes like color, fabric, style, or occasion. For instance, someone can take an image of a blazer and adjust the search to find more formal versions of that item or to explore variations with different aesthetics like “Y2K.”

The company is also expanding where visual search can be accessed within the app. Users can now initiate a search by long-pressing on any Pin in their Home Feed, which triggers a visual search without requiring typed input.

These updates are being introduced shortly after Pinterest announced it would begin labeling AI-generated images and offer users the option to limit how much of that content appears in their feeds. This move follows user concerns about the increasing presence of AI-generated images diluting the platform’s value for discovering real products and ideas.

The updated visual search tools reflect Pinterest’s ongoing shift toward supporting image-based navigation and discovery, with an emphasis on enabling users to interact more directly with the platform’s visual content.