ChatGPT MarkItDown MCP

A free, read-only Microsoft MarkItDown MCP that augments ChatGPT's native file handling. ChatGPT remains the reasoner; this service provides semantic document extraction where that adds value.

Production

• MCP: https://chatgpt-markitdown-mcp-ten.vercel.app/mcp

• Health: https://chatgpt-markitdown-mcp-ten.vercel.app/health

• Production source: GitHub main

Related MCP server: mcp-server-convert

Tools

• extract_file — accepts a normal ChatGPT attachment through native openai/fileParams, fetches it through the hardened downloader, converts it in memory, and returns bounded Markdown with continuation metadata.

• convert_to_markdown — converts public HTTPS resources and small data URIs through the same network safety boundary.

• markitdown_capabilities — describes supported formats and privacy limitations.

Format strategy

MarkItDown supports the free built-in PDF, DOCX, PPTX, XLS/XLSX, Outlook, HTML/text, CSV/JSON/XML, ZIP, EPUB, audio, YouTube, and related conversion paths installed by this project. Images are intentionally better handled by ChatGPT's native vision unless a future tested visual-PDF fallback is enabled.

The server does not enable paid Azure/OpenAI OCR/document services.

Privacy note for audio

MarkItDown 0.1.7's built-in speech transcription calls Google Speech Recognition. For that reason, private attached audio/video is blocked by default and requires explicit user consent before the tool enables that path.

Security

See SECURITY.md. Key controls include SSRF protection with validated/pinned public IPs, 5-second connect and 10-second read timeouts within a 30-second total fetch deadline, three redirects, a 32 MiB download ceiling, archive preflight, 50,000-character output chunks, no file persistence, minimized logging, and explicit prompt-injection boundaries.

Development verification

python -m pytest -q
python -m py_compile server.py document_intelligence.py

Vercel production deploys from main; do not create out-of-band production drift.