▲ 259 today192k starsPythonHN ×2
Microsoft's MarkItDown turns PDFs, slides, and audio into LLM-ready Markdown
What is markitdown?
MarkItDown is a Python utility that extracts text and structure from PDFs, Word docs, PowerPoint decks, spreadsheets, images, audio, HTML, and more, then outputs it as Markdown. The result is token-efficient and suited for prompting large language models or downstream text analysis.
Who it is for
If you build LLM pipelines or data-science workflows, you can ingest a wide range of document types without writing custom parsers.