MarkItDown

MarkItDown is an online converter inspired by Microsoft's open-source MarkItDown utility. It handles PDF, DOCX, XLS/XLSX, PPTX, HTML/HTM, CSV/TSV, JSON, XML, TXT/MD, EPUB, IPYNB, public URLs, and PNG/JPG/WebP images via OCR. Users can choose local browser processing for supported files or server conversion for richer parsing. The site states that the output can be previewed, edited in the browser, copied, downloaded as .md, or saved when signed in. The free tier allows up to 3 files per batch and files up to 20MB, with Pro raising limits; no account is needed to start. The page states that the original project is MIT-licensed and that the site self-hosts the Microsoft MarkItDown service.

Key features

• Converts PDF, DOCX, PPTX, XLS/XLSX, HTML/HTM, CSV/TSV, JSON, XML, TXT/MD, EPUB, and IPYNB files to Markdown
• Converts public URLs and pasted tables to Markdown
• Local OCR and AI OCR for PNG/JPG/WebP images and scanned PDFs
• Preserves tables, headings, lists, links, and inline formatting where the parser can read them
• Choice between local browser processing and server conversion
• Preview raw output including tables, code, LaTeX, and Mermaid before use
• Browser editor, copy to clipboard, and .md download
• Optional saving of results when signed in
• Batch processing (signed-in and Pro tiers)

Best suited for

• Preparing documents for LLM prompts
• Feeding RAG pipelines, vector databases, and knowledge bases
• Converting documents for search indexes, docs, and Git workflows
• Extracting tables and structure from scanned documents with OCR

Pricing

Free to start without an account (up to 3 files per batch, 20MB per file, limited OCR pages). Pro tier is referenced with higher limits; specific Pro prices are not stated on the page.