opendataloader-pdf
PDF Parser for AI-ready data. Automate PDF accessibility. Open-source.
What is opendataloader-pdf?
PDF Parser for AI-ready data. Automate PDF accessibility. Open-source.
Overview of opendataloader-pdf
What it is
OpenDataLoader PDF is an open-source PDF parser for AI-ready data extraction and PDF accessibility automation. It extracts Markdown, JSON (with bounding boxes), and HTML from PDFs. It offers a deterministic local mode and an AI hybrid mode for complex pages. It also provides auto-tagging to convert untagged PDFs into screen-reader-ready Tagged PDFs.
Key features
- Extracts text, tables, images, formulas, and headings with correct reading order and bounding boxes - Hybrid mode combines deterministic local parsing with AI backend for complex pages (scanned PDFs, borderless tables, charts, formulas)
- Built-in OCR supporting 80+ languages in hybrid mode - Auto-tagging of untagged PDFs into Tagged PDFs (Apache 2.0, no proprietary SDK dependency)
- Outputs structured Markdown, JSON with bounding boxes, and HTML - Ranks #1 in benchmarks with overall accuracy 0.907, table accuracy 0.928 - Available SDKs for Python, Node.js, Java
Use cases
- AI data extraction from PDFs for RAG pipelines, document analysis, and structured data generation - PDF accessibility compliance (EAA, ADA, Section 508) by auto-tagging untagged PDFs - Processing complex PDFs including multi-column layouts, scanned documents, mathematical formulas, and charts - Batch conversion of PDFs to Markdown/JSON for downstream chunking and indexing
Getting started
bash pip install -U opendataloader-pdf python import opendataloader_pdf opendataloader_pdf.convert( input_path=[“file1.pdf”, “file2.pdf”, “folder/”], output_dir=“output/”, format=“markdown,json” )
License and pricing
License: Apache-2.0. The project is open-source. Auto-tagging is free under Apache 2.0. Enterprise add-ons (PDF/UA-1/2 export, accessibility studio) are available separately.
Based on the project's public README and repository metadata.