Finds duplicate, modified and even derivative images in your image collection.

Your own private search by image API.
Fast, scalable and accurate.

MatchEngine works with your own image collection and finds duplicate, resized and modified images. Powered by TinEye's unparalleled image recognition, MatchEngine is engineered to deal with a broad range of image transformations, including resizing, cropping, edits, occlusions and color changes, amongst others.

MatchEngine is the API of choice for: identifying duplicate images, profile and UGC image verification, fraud detection, image collection reconciliation, and blacklisting unwanted images.

MatchEngine is scalable and once integrated enables you to include image recognition capabilities in your own applications, enterprise solutions and web services.

Powering millions of searches

MatchEngine works with your own large scale reference image collection

Quickly identify duplicate and altered images in any size image collection

Easily reconcile large image collections against each other

Eliminate your reliance on metadata for image verification and compliance

Compare all incoming profile or UGC images against your image blacklist

Scalable, high performance features for performing reverse image searches

Identify and group duplicate product images, despite watermarks, crops and many other image transformations

How it works

MatchEngine uses image recognition to index your image collection. MatchEngine creates a unique fingerprint for each image in your collection and does not use any data associated with your images to create the image recognition based fingerprint.

These unique digital fingerprints describe the patterns found in the pixels of the images. MatchEngine is able to identify even partial matches of these fingerprints.

MatchEngine creates an index of all your images’ fingerprints for reference and stores them in an internal database. API calls are used to add images to the index and to search through existing images. Image matching is quick, and scales to handle hundreds of millions of images.

Hosted API

MatchEngine is a fully hosted API. There is no software to install and no updates to worry about. By using our hosted API, you save time and money needed to build and maintain your own image recognition solutions.

Easy to integrate

MatchEngine is easily integrated with your existing technology, regardless of the languages and tools you use.

Fully documented

The MatchEngine API is a REST-based API. REST/JSON bindings are available in all major programming languages. We also support a growing number of language-specific libraries to make integration very straight forward. Our documentation and libraries are extensive.

Scalable and high performance

MatchEngine scales. It can index anywhere from thousands up to hundreds of millions of images. Searching is fast, and optimized for large scale image collections.

Proven technology

MatchEngine is based on TinEye's highly scalable and accurate image recognition technologies. Built by TinEye's image recognition scientists and backed by a decade of research and development, MatchEngine is the ideal image recognition solution for image verification.

MatchEngine: Image recognition customized for you.

Works with your image collection.


  • up to 5,000 images
  • 1,000 monthly searches
  • $200
  • USD per month
Sign up
Need more?
  • $0.01 per image per month
  • $0.01 per search


  • up to 50,000 images
  • 30,000 monthly searches
  • $500
  • USD per month
Sign up
Need more?
  • $0.005 per image per month
  • $0.005 per search


  • up to 500,000 images
  • 150,000 monthly searches
  • $1500
  • USD per month
Sign up
Need more?
  • $0.003 per image per month
  • $0.003 per search


  • up to 500 million images
  • 50 million monthly searches
Contact Us
Custom plans
  • We customize plans for large collections
  • that will scale to your needs.

Trusted & used by

Brand View

Frequently asked questions

Q.How does MatchEngine work?

MatchEngine uses image recognition to search images. Once you submit your images to MatchEngine, they are automatically fingerprinted using our image recognition and indexed for searching. The unique fingerprint generated for each image is what MatchEngine uses to distinguish one image from another.

  • A unique digital fingerprint is extracted from each image based on the patterns within the image.
  • Fingerprints for your images are stored in an index for your entire image collection.
  • When searching for an image, a fingerprint is generated on the fly and compared against your complete fingerprint index to identify matches.
  • Image submissions, via the 'add' API method, will update your index automatically.
Q.How are my images stored?

Your images are stored on our secure datacenter servers. Access to your image collection is handled by the MatchEngine API, and only you have access to your images.

Q.What kind of images does MatchEngine recognize?

MatchEngine can match images that have been altered: resizing; cropping; Photoshop edits; occlusions; color changes; etc. However, it is not optimized to work with the 'noisy' photos taken by mobile phones – if this is your use case please see MobileEngine.

Q.How do I add my images to the MatchEngine API?

Images are submitted to the MatchEngine API via an HTTP POST using the API‘s 'add' function. Submitted images are copied to our servers where they are automatically fingerprinted using our image recognition technology and made available for searching. The unique fingerprint generated for each image is what the MatchEngine uses to distinguish one image from another and identify matches. Images should be prepared as follows:

  • JPEG, PNG, GIF, BMP, or TIFF format
  • Scaled to ~600 pixels in the smallest dimension

MatchEngine will automatically scale larger images down to 600 pixels. However, large images increase transmission and processing time.

Q.Can I access MatchEngine over HTTPS?

All of our APIs, including MatchEngine, can be accessed over HTTPS to keep your data protected.

Q.How do I know if MatchEngine API is right for me?

Get in touch and tell us about your search challenge or project. We will be happy to work with you to identify the best API for your needs. Common uses for MatchEngine include: de-duplicating large image collections, searching incoming images against either a black or white list, verifying profile or user generated images, and clustering images.

Q.Where can I find documentation about the MatchEngine API?

Our complete MatchEngine documentation can be found here. This also includes sample code, supported libraries, calls supported by the API, tutorials and more.