The best document parsers, according to hackers
8 document parsers recommended in Hacker News comments over the past year, ranked by how many separate people recommend each one.
- Most recommendedPandoc142 people recommend it
- Least regretMozilla Readability10 recommend, 1 criticize
All document parsers
- PandocThe go-to tool for converting documents between nearly any formats, extensible with filters and templates; critics cite a huge binary, slow speed, and fiddly setup.142recommend
- Tesseract OCRFree open-source OCR engine, fast and accurate on clean text across many languages. Critics say it struggles with complex layouts and handwriting.32recommend
- DoclingFree open-source tool that turns PDFs and tables into markdown and text with traceable positions. Complex tables and multi-column layouts can lose structure, and it is slow.19recommend
- Mozilla ReadabilityFree open-source library that strips clutter from web pages to leave the article text, and is built into some browsers. Caveat: it does not always work on paywalled pages.10recommend
- ZpdfFast PDF text extraction library in Zig, reported several times faster than MuPDF and MIT licensed. Critics say it is vibe coded, untested, and crashes on some files.8recommend
- MarkItDown by MicrosoftPeople recommend it for converting documents to Markdown, especially as clean raw text for LLM agents, and say it works well most of the time.7recommend
- AsciidoctorMature, free tool for writing documentation and books in AsciiDoc, with many output formats. Some find it involves compromises and a Ruby dependency.7recommend
- Azure Document IntelligencePeople recommend it for accurate OCR, including handwriting, and for extracting complex tables, with a free tier for small volumes. One commenter warned that its pricing can be expensive and surprising.5recommend