Yandex Cloud Vision OCR icon

Yandex Cloud Vision OCR

Recognize text in images using Yandex Cloud Vision OCR API

Actions3

Text Recognition → Recognize

AI-generated

Summary

Recognize text synchronously in image or PDF binary data using Yandex Cloud Vision OCR with configurable OCR models, languages, and output formats.

Inputs

  • Binary Property (required) — The name of the binary property in the input containing the image or PDF data to process.
  • MIME Type — MIME type of the input image or PDF; can be set to auto-detect or manually selected (JPEG, PNG, PDF). Defaults to auto-detect.
  • Languages — One or more languages to recognize in the image, defaulting to Russian and English.
  • Model — The OCR model to use, e.g., general page text, multi-column, handwritten, tables, markdown, passport, license plates, and document templates.
  • Output Format — Format of returned data: full recognized text string, detailed structured data, or both. Defaults to full text only.

Output shape

A list of JSON objects each containing recognized text results for an input binary item.

Each output item provides a JSON with fields depending on the selected output format: 'fullText' as a string, 'structured' as detailed blocks/lines/words and coordinates, or both combined. Errors related to missing binary data or size limits (10MB max) for synchronous recognition will be thrown per item.

Links

Discussion