Lenso

Lenso (lenso.ai) is an AI-powered reverse image search engine that starts a search with a picture instead of keywords. The platform is built on CBIR — content-based image retrieval — and runs several AI-trained models, one per result category, so a single upload can be explored from different angles: People, Places, Duplicates, Similar, and Related. Each category produces its own set of matches rather than one blended list, drawn from an index of billions of images crawled across the web.

The People category is a facial recognition search. Lenso measures distances between facial features such as eyes, nose, and mouth to locate the same face in different photos, poses, and expressions. It is aimed at people who want to see where their own photos surface online, and it only runs after explicit consent because the image is sent to EyeMatch AI PTE. LTD. for processing. Facial search is available only in selected regions.

The Duplicates category targets exact copies and edited variants — cropped, filtered, or resized versions — and returns the source pages hosting each copy, which makes it useful for photographers and rights holders chasing unlicensed reuse. Similar returns visually alike images with comparable layout or content, Related returns images connected to the subject without looking alike, and Places identifies landscapes, buildings, and landmarks. Every match arrives attached to the page it sits on, so the practical output of a search is a list of URLs rather than a gallery — which is why unlocking those links, not running the search, is the paid step.

Beyond the search itself, Lenso adds workflow tools. Results can be sorted by best match, newest, or most diverse, filtered by keyword or domain, and shared through a copyable link that stays private to you. Alerts watch a category and email you when new matches appear, and Collections group unlocked results by topic. The Professional plan unlocks Research Mode, which reaches up to 10,000 results for People and Duplicates — roughly ten times the normal result set — plus Advanced Filters for language, date range, domain, image size, and title keyword. It also adds DMCA Assist, which collects the image links and sources you pick and generates a ready-to-copy takedown request for the hosting company.

A Developer plan exposes the same engine as an API: the Image Search API covers Duplicates, Places, Related, and Similar, while the Face Search API mirrors the People category, with up to 5,000 requests per month and usage statistics in the user panel. Uploads accept JPG, PNG, and WEBP up to 10MB and at least 200x200 pixels, the interface is localised into sixteen languages, and a downloadable Chrome extension puts the upload field in the browser toolbar. An opt-out request form also lets anyone ask to have their images excluded from the index.