Parses PDFs and scanned pages, extracting structured text, tables and formulas
Overview
- Purpose
- Parses PDFs and scanned pages, extracting structured text, tables and formulas
- Category
- Office
- Developer
- OpenDataLab
- Official site
- https://mineru.net/ (opens in a new tab)
Main features
- Parses PDFs, images and some Office documents
- Recognizes scanned pages and extracts text
- Recognizes document structure such as tables, images and formulas
- Outputs Markdown, JSON and other formats
Best for
Suited to turning papers, scanned material or documents with complex layouts into searchable, editable content, such as organizing formulas, tables and knowledge base material.
Things to know
- Conditions differ between the online version and local deployment; local installation needs a suitable environment and resources
- Scan quality, complex layouts and handwriting affect recognition
- Check formulas, tables and paragraph order against the original
FAQ
- Can it handle scanned PDFs?
- The official project states it recognizes scanned PDFs with OCR enabled; results depend on the quality of the original.
- Can it export Markdown?
- Yes. The official project lists Markdown, JSON and other output formats.