OpenDataLoader PDF
PDF parser that outputs Markdown, JSON and HTML, plus PDF auto-tagging
What it is
A Java-based PDF parser with Python, Node.js and Java SDKs that converts PDFs into Markdown, HTML, and JSON with bounding boxes, reading order, tables, headings and lists. A hybrid mode sends hard pages to an AI backend for OCR on scans, borderless tables, LaTeX formulas and chart descriptions. It can also add accessibility tags to PDFs that lack them, producing Tagged PDFs that screen readers can navigate. Developers building RAG pipelines use it to prepare documents for chunking and citation, and organizations use it to cut the cost of accessibility remediation.