TinEye

2008

Public launch

TinEye blog

$0

Web search

tineye.com

$0.02

API search at the 50,000 bundle

TinEye blog, bundle pricing

What does TinEye do?

TinEye answers one question: where else has this exact picture been published? It computes a fingerprint of the uploaded image and compares it with fingerprints of images it has crawled. Matches include the same file, resized and recompressed copies, crops, colour-adjusted versions and pictures with text or watermarks added. TinEye's own description is that it uses image recognition and a growing index of billions of images, and that it does not look at keywords or page text (TinEye blog, checked November 6, 2026).

It is the reference example of a reverse image search rather than a reverse face search: a portrait finds reposts of that portrait and nothing else of the person. For a face, TinEye returns no results unless the same photo exists somewhere, which makes it useful for checking whether a dating-profile picture was lifted from elsewhere and useless for finding a person's other photos.

How do you use it?

1. Step 1

   Submit an image

   Upload a file, paste an image URL, drag an image onto the page, or right-click an image with the browser extension.

   ›
2. Step 2

   Fingerprint

   TinEye computes a compact signature of the picture; it says the image is not stored after the search.

   ›
3. Step 3

   Match

   The signature is compared with the index; results list every page with a matching copy.

   ›
4. Step 4

   Sort and filter

   Sort by best match, most changed, biggest image, newest or oldest; filter by domain or collection.

What are the sort options for?

• Best match puts the closest copies first, usually the original or the highest-quality copy.
• Most changed surfaces edited versions, which is where added text, watermarks and composites show up.
• Biggest image is the fastest way to find the original resolution.
• Newest and oldest order by the date TinEye first crawled each copy, which is often the best available clue to where a picture started.

What does it not do?

Those limits are why investigators and journalists pair TinEye with a face-first engine: TinEye to find the origin and spread of a particular photo, a face search such as FaceSearch or PimEyes to find other photos of the person. The concepts section explains the technical difference.

What happens to an uploaded image?

TinEye states that images submitted for search are used to generate the fingerprint and are not added to the index or kept, and that searches are private. Its FAQ and help centre are the reference for retention, file formats (JPEG, PNG, GIF, BMP, TIFF, WebP) and the size limit, which is 20 MB per image per the help centre. Those pages block automated fetching, so the figures here were taken from the pages as rendered in a browser on November 6, 2026.