# PaddlePaddle/PaddleOCR

Turn any PDF or image document into structured data for your AI. A powerful, lightweight OCR toolkit that bridges the gap between images/PDFs and LLMs. Supports 100+ languages.

Repository: https://github.com/PaddlePaddle/PaddleOCR
Canonical: https://ross.abutalabs.com/products/paddleocr
Homepage: https://www.paddleocr.com
Language: Python
License: Apache-2.0
License Family: permissive
Topics: ocr, chineseocr, pdf2markdown, pp-ocr, pp-structure, document-parsing, document-translation, kie, ai4science, pdf-extractor-rag, pdf-parser, rag, paddleocr-vl
Last push: 2026-07-22T11:59:34+00:00

## Health v2 (maintenance only)
Score: 93/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 93, release rhythm 88, longevity 100
- inputs: {"age_days": 2308, "days_push": 42, "days_rel": 83, "gap_med": 14.5, "n_releases_24m": 19}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 88312, forks 11237 (observed 2026-08-28T04:12:23.085696+00:00)

## What it is
PaddleOCR is a multilingual OCR and document parsing toolkit built on PaddlePaddle that converts images and PDFs into structured data like Markdown. It supports 100+ languages, text detection/recognition, layout analysis, and key information extraction, and is designed to feed documents into LLM and RAG pipelines.

## Use cases
- extract text from images and scanned pdfs
- convert pdf documents to markdown for llm ingestion
- build a rag pipeline over document collections
- recognize chinese and other multilingual text in photos
- parse invoices and forms to extract key fields
- ocr tables and document layout into structured data

## When to choose
- you need high-accuracy multilingual ocr, especially for chinese text
- you want to turn pdfs or images into structured input for llms or rag
- you need document parsing with layout, table, and key-information extraction
- you want a mature, actively maintained toolkit with cpu and gpu support

## When to avoid
- you need a lightweight pure-inference solution without a deep learning framework dependency
- your project is built on pytorch or tensorflow and you don't want paddlepaddle
- you only need simple tesseract-level ocr without document structure parsing

## Facets
- artifact type: library
- maturity: stable
- function: ocr, pdf, machine-learning, rag, nlp, image-processing
- domain: computer-vision, pdf, artificial-intelligence
- platform: python, windows, cross-platform
- tags: document-parsing, text-recognition, text-detection, multilingual, paddlepaddle, pdf-to-markdown, key-information-extraction, document-ai, natural-language-processing, retrieval-augmented-generation, linux, macos, gpu

## Member repositories
- PaddlePaddle/PaddleOCR (main) score 93

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:12:23.085696+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T16:13:52.999194+00:00, confidence not recorded.
  - readme: https://github.com/PaddlePaddle/PaddleOCR (fetched 2026-08-28T04:12:23.085696+00:00, sha e37865b4d7d5)
  - homepage: https://www.paddleocr.com (fetched 2026-08-28T17:38:44.604727+00:00, sha e962d88d5568)
  - registry_pypi: https://pypi.org/pypi/paddleocr/json (fetched 2026-08-28T17:38:44.613238+00:00, sha 405135b3572d)
- Data as of 2026-08-30T08:39:29.467469+00:00.