Download Overview - Real-Time Recognition SDK

Real-Time Recognition SDK is a software development kit that enables developers to create their own mobile apps with included functionality of recognizing text directly on smartphone camera preview screen.

Latest Release

  • Real-Time Recognition SDK 1 - Release 4
    • Release Date: 27.06.2017
    • Part #: 1305/18
    • Build #:

Download the latest release:

New Features and Improvements

  • Ability to store patterns and dictionaries outside of the application's package/bundle
    This feature allows reducing the size of the app. Additional patterns and dictionaries can be offered separately: through the app store or via in-app purchase.
  • [Technical preview] Data Capture support
    This release contains new interfaces, which can be used by developers to implement field extraction (Data Capture) in their apps. This feature is available as a technical preview.
  • iOS 7 support discontinued.
    Starting from this release ABBYY Real-Time Recognition SDK no longer supports iOS 7.

Code Samples

  • Once you have downloaded Real-Time Recognition SDK, feel free to check the code sample on GitHub:

Further Information

Request Trial

  • Note: Just to save your time: SPAM requests will be deleted right away - Read here why...

Previous Releases

Release 3

Real-Time Recognition SDK 1 - Release 3

  • Release Date: 02.05.2017
  • Part #: 1305/17
  • Build #:

New Features and Improvements

  • Support of ARM64 Android processors
    The Android library version compiled for ARM64 processors is added to the distribution kit.
  • Recognition of Russian Passport
    API is extended and new interfaces are added supporting implementation of Data Capture scenarios. The current release supports one new scenario, enabling these interfaces for recognition of Russian passport. The distribution kit contains all necessary data and a sample code demonstrating how to capture passport using Real-Time Recognition SDK.
  • Recognition of a Single Photo
    API is extended, adding a capability to process a single image (not only video stream). It is useful for scenarios, when the document to be recognized is sent via e-mail.
  • Working with Custom Dictionaries
    Now Real-Time Recognition SDK supports custom dictionaries (similar to Mobile OCR Engine). To recognize using custom dictionaries developer needs to have special pattern and dictionary files, which he/she can get from ABBYY technical support.

Release 2

Real-Time Recognition SDK 1 - Release 2

  • Release Date: 14.03.2017
  • Part #: 1305/16
  • Build #:

New Features and Improvements

  • Improved text analysis
    The algorithm conducting image analysis and detecting text areas on video stream frames is now more stable and more accurate.
  • Improved error handling
    The error message helps developers to notice an error and apply a corrective measure.

Release 1

Real-Time Recognition SDK - Release 1 Update

  • Release Date: 15.12.2016
  • Part #: 1305/15
  • Build #:


  • Real-Time OCR
    Thanks to Real-Time Recognition SDK, developers may create mobile apps with the capability to recognize text directly on the smartphone's camera preview screen. In comparison with ordinary OCR technologies that work with photos, Real-Time Recognition SDK provides higher accuracy and doesn’t require snapping and saving a photo, which improves customer experience, and ensure confidentiality.
  • Recognition of text in real-world scenes
    Real-Time Recognition SDK is able to find text of any color on any background (like road signs and menus) and extract it.
  • Support of 63 recognition languages
    Using Real-Time Recognition SDK it is possible to recognize text in up to 63 languages including Chinese, Japanese, Korean and several languages written in Cyrillic characters.
  • Text translation
    The extracted text can be translated instantly on smartphone camera preview screen. The new SDK provides build-in translation dictionaries for several languages.
  • Works on device
    All of the processing is performed offline on the mobile device, so the mobile application with Real-Time Recognition SDK can work without internet connection.
  • Code Samples for text capture
    The distributive includes code samples for text capture scenario in Swift and Objective C for iOS as well as in Java for Android. The samples demonstrate how to work with the library to implement text capture scenario. The Samples are an excellent basis for a user to get quickly started with the development.
  • iOS and Android support
    Real-Time Recognition SDK is available for platforms iOS and Android.