All open source
Open SourceScraping60k

MarkItDown

Convert documents and web content into clean LLM-ready Markdown.

MarkItDown is a Microsoft utility that converts a wide range of files and web pages (HTML, PDF, Office docs, images, audio) into clean Markdown optimized for LLM consumption. It preserves structure like headings, tables, and lists so models can reason over content reliably. It is notable for normalizing messy scraped or downloaded content into a single text format for ingestion.

Repository

microsoft/markitdown

Language

Python

What you'd build with it

  • Turning scraped HTML pages into clean Markdown for RAG ingestion
  • Normalizing PDFs and Office docs into LLM-friendly text
  • Preprocessing mixed web content into a consistent format for embeddings

Tags

markdownllmconversionrag