Opendataloader Pdf

OpenDataLoader PDF is an open source parser that converts PDF files into structured, LLM ready output. It produces Markdown and JSON, and every element in the JSON carries a bounding box, page number, heading level and font details, so a retrieval augmented generation pipeline can cite the exact location of a passage in the source document. It targets developers building RAG and document pipelines who find that standard PDF text extraction scrambles multi column layouts and flattens tables. The parser restores reading order with an XY Cut algorithm, detects table borders and merged cells, and handles formulas, pictures and charts. OCR supports more than 80 languages. An optional hybrid mode adds an AI model for scanned pages and complex tables. A safety layer filters hidden text, off page content and prompt injection attempts before text reaches a model. The project also offers a second use: auto tagging an untagged PDF into a Tagged PDF so screen readers can use it, based on PDF Association specifications and validated with veraPDF, which supports accessibility compliance work. Processing runs fully locally. The project provides quick starts for Python, Java and Node.js, a command line, and a live demo. It is released under the Apache 2.0 license. The site publishes no paid plans or prices.