---
id: "b927"
language: "en"
library: "tools"
official_url: "https://github.com/PaddlePaddle/PaddleOCR"
profile_updated_on: "2026-10-23"
---

# PaddleOCR

OCR, layout and table parsing for developers converting images and PDFs into structured content.

- Company / organization: PaddlePaddle project
- Category: Model development
- Official website: [https://github.com/PaddlePaddle/PaddleOCR](https://github.com/PaddlePaddle/PaddleOCR)
- Profile updated: 2026-10-23


## Overview

PaddleOCR is a toolkit for document AI. PP-OCR detects and recognizes text, PP-Structure extracts layouts and tables, and PaddleOCR-VL offers a vision-language parsing path. Choose pipelines for OCR, coordinates or Markdown/JSON before integrating retrieval or agents. Language coverage and output granularity vary by model family; toolkit-wide claims do not apply to every model.

## Usage guide & costs

### Before you start

Code is Apache 2.0; configure runtimes, weights and inference backends separately. A model’s language count or speed does not describe all pipelines, and the toolkit is not a free hosted API.

[Official details](https://github.com/PaddlePaddle/PaddleOCR/blob/main/README.md)
[Official details](https://github.com/PaddlePaddle/PaddleOCR/blob/main/LICENSE)


## Official sources and references

- [Official website](https://github.com/PaddlePaddle/PaddleOCR)
- [Documentation](https://www.paddleocr.ai/latest/)
- [Documentation](https://github.com/PaddlePaddle/PaddleOCR/blob/main/README.md)

## Reading & subscriptions

- [Interactive profile](../../../products/paddleocr--b927.html)
- [Read](../../../read/en/products/paddleocr--b927.html)
- [JSON](../../../read/en/products/paddleocr--b927.json)

Descriptions summarize public information; consult official websites for features, pricing and regional availability.