Why Markdown first
Markdown is easier for LLMs, retrieval pipelines, and human review than raw binary documents or messy copy-pasted PDF text.
LLM and RAG
PDF to Markdown for LLMs
Prepare PDF content for LLM and retrieval workflows with Markdown output.
v1 converter path
Upload -> Convert -> Preview -> Copy -> Download
Supported: PDF, Office, HTML, CSV, JSON, XML, TXT, MD
Limit: one local file, up to 10 MB
Deferred: OCR, batch mode, API keys, full MCP server
RAG workflow fit
Use the converter before chunking, embedding, summarization, or manual cleanup. v1 keeps the output visible and downloadable.
Safety boundary
The service does not store user files permanently and does not send content to an LLM provider in v1.
Trust and safety
This site is not affiliated with Microsoft. MarkItDown is an open-source project associated with Microsoft; this website is an independent tool-site experiment. v1 uses temporary file conversion and does not store files permanently.
FAQ
Does the converter call an LLM?
No. v1 uses document conversion, not a paid AI model call.
Is it suitable for scanned documents?
Not yet. Scanned PDF OCR is a later version item.