Yandex Vision

A computer vision service for image analysis.

Support for 40+ languages
The service uses different language models. These include the Latin-Cyrillic model that supports more than 40 languages.
Easy integration
The service provides the REST and gRPC APIs that you can use to integrate it with your applications.
High-speed processing
It takes a couple of seconds to recognize one page of text. Each request can send multiple images containing text in one or more languages.
Image classification
Preview
Image classification technology helps analyze the content that your users upload. For example, you can filter adult content or select only high-quality images.
Text recognition
Yandex Vision recognizes text in an image, specifies its location, and returns it as blocks of text, lines, and words with automatic punctuation.
Face detection
Preview
Face detection automatically finds photos with portraits in all images and helps mark people in a photo.
License plate number recognition
The service recognizes license plate numbers in photos and document templates, such as passports.

Implement your projects using Yandex Vision

Search for text in an archive of scanned documents

If you need to find specific information in an archive of scanned documents that are only saved as images, Yandex Vision helps you recognize text and makes it available for indexing and searching.

Automatic translation of text in images

Recognize text in an image and immediately translate it to other languages using Yandex Vision and Yandex Translate. To embed this feature in your apps, use the REST or gRPC API that both services provide.
Works with
Translate

Automate your workflow

Use Yandex Vision to automate your business processes where you need to scan and recognize text in standard types of documents and add them to your document workflow system. The service recognizes text and analyzes which part of a document it should be placed in.

Questions and answers

What types of analysis does Yandex Vision support?

Yandex Vision supports the following features: optical character recognition (OCR), document recognition, license plate number recognition, face detection, and image classification.

Yandex Vision supports the following features: optical character recognition (OCR), document recognition, license plate number recognition, face detection, and image classification.

What models are used for classification?

The service provides image classification based on a model that is trained on a specific set of images. Currently supported models:

  • quality: Assesses the quality of an image.
  • moderation: Assesses whether an image meets moderation requirements.

The service provides image classification based on a model that is trained on a specific set of images. Currently supported models:

  • quality: Assesses the quality of an image.
  • moderation: Assesses whether an image meets moderation requirements.

What file formats does Yandex Vision support?

Yandex Vision works with JPEG, PNG, and PDF files.

Yandex Vision works with JPEG, PNG, and PDF files.

Get started with Yandex Vision