MCP Servers / file-system / pdfmux

pdfmux

JSON →
stdionone65file-system

Smart PDF-to-Markdown router that picks the best extractor per page, audits output quality, and re-extracts failures automatically. Confidence scoring, BYOK LLM support, RAG chunking.

Install
How to run this server
[ { "cmd": "npx -y pdfmux-mcp", "imports": [] } ]
server path: pdfmux
Tools
What this server exposes
pdfmux convert
Extract PDF to Markdown, JSON, or chunks with per-page confidence scoring, auto-routing each page to the best backend.
pdfmux stream
Stream pages as NDJSON as they finish, useful for long documents.
pdfmux watch
Watch a directory for new PDFs and auto-convert them.
pdfmux estimate
Predict cost before running extraction on a PDF.
pdfmux diff
Diff two extractions side-by-side.
pdfmux doctor
Pre-flight a directory to check which extras are needed for the batch.
batch_extract
Batch extract PDFs, yielding (path, result) tuples as each completes.
extract_text
Extract PDF to a Markdown string.
extract_json
Extract PDF to a locked schema dict.
chunk
Extract PDF into RAG-ready chunks with token limits.
Configuration
Environment & auth
authnone
envGEMINI_API_KEY
envANTHROPIC_API_KEY
Resources
pdfmux — MCP Server · libregistry