PDF MCP servers for reading and creating documents
MCP servers that read, search, split, fill, and generate PDF files, including OCR for scans, so AI assistants can work with real documents.
What to know about PDF MCP servers
PDF servers solve a problem that chat clients handle unevenly: getting reliable text, tables, and layout out of a PDF, or producing a new one. Reading servers extract text page by page, keep track of page numbers for citations, and in better cases preserve tables. Writing servers fill forms, merge or split files, and render Markdown or HTML into a finished PDF.
Scanned documents need OCR, and that is where servers differ most. Some run Tesseract locally, some call a cloud document AI service, and some hand page images to a vision model. Local OCR is private and free; cloud services are more accurate on handwriting and complex layouts.
For long reports and contracts, prefer servers that let the agent request specific pages or search within the document. Loading a three hundred page PDF into context at once is slow and often truncated.
PDF MCP servers
The 12 strongest of 1001 matching servers. Servers that name the topic directly come first, then by trust score.
- 72Trust score 72 of 100, Good
Agentic RAG over one PDF or a whole folder: hybrid search, selective page reads, tables, OCR.
RegistryLicenseQualityMaintenanceLicense: A. MIT is a permissive open source license.Quality: Not graded. No README or tool list has been captured yet.Maintenance: A. Last commit or release 10 days ago. - 71Trust score 71 of 100, Good
PDF Reader (PDF Agent Stack)
by shuji-bonji · TypeScript
Read PDF internals: text, structure, fonts, signatures and tags, with what was not read declared.
RegistryLicenseQualityMaintenanceLicense: A. MIT is a permissive open source license.Quality: Not graded. No README or tool list has been captured yet.Maintenance: A. Last commit or release 23 days ago. - 71Trust score 71 of 100, Good
PDF MCP server: PDF/A, PDF/X-4, PAdES (LTV), forms, 32 scripts, JPEG 2000, spot colours. 28 tools
RegistryLicenseQualityMaintenanceLicense: A. MIT is a permissive open source license.Quality: Not graded. No README or tool list has been captured yet.Maintenance: A. Last commit or release 17 days ago. - 68Trust score 68 of 100, Good
PDF Verify (PDF Agent Stack)
by shuji-bonji · TypeScript
Verify PDF signatures, detect tampering, and identify PDF/A and PAdES conformance levels.
RegistryLicenseQualityMaintenanceLicense: A. MIT is a permissive open source license.Quality: A. 93/100 for listing completeness. Lost points: 1 of 8 tools barely described, short summary.Maintenance: A. Last commit or release 25 days ago. - 67Trust score 67 of 100, Good
PDF Spec (PDF Agent Stack)
by shuji-bonji · TypeScript
Look up ISO 32000 (PDF) clauses, requirements and terms. A reference, not a conformance checker.
RegistryLicenseQualityMaintenanceLicense: A. MIT is a permissive open source license.Quality: C. 63/100 for listing completeness. Lost points: no README, 4 of 12 tools barely described, short summary.Maintenance: A. Last commit or release 13 days ago. - 65Trust score 65 of 100, Good
PDF Writer (PDF Agent Stack)
by shuji-bonji · TypeScript
Create and edit PDFs: text, Markdown, tables, metadata and pages, with Japanese font embedding.
RegistryLicenseQualityMaintenanceLicense: A. MIT is a permissive open source license.Quality: Not graded. No README or tool list has been captured yet.Maintenance: A. Last commit or release 25 days ago. - 58Trust score 58 of 100, Fair
Compile LaTeX to PDF without installing TeX Live. No API key needed to start.
RegistryRemoteLicenseQualityMaintenanceLicense: Not graded. No license is recorded for this server.Quality: Not graded. No README or tool list has been captured yet.Maintenance: Not graded. No commit or release date has been recorded. - 55Trust score 55 of 100, Fair
Create, preview, inspect, verify, and compile deterministic PDF documents with Fullbleed.
RegistryLicenseQualityMaintenanceLicense: Not graded. No license is recorded for this server.Quality: A. 91/100 for listing completeness. Lost points: short summary.Maintenance: Not graded. No commit or release date has been recorded. - 55Trust score 55 of 100, Fair
Extract local PDF text with page-level paragraphs and English/German OCR fallback.
RegistryLicenseQualityMaintenanceLicense: Not graded. No license is recorded for this server.Quality: Not graded. No README or tool list has been captured yet.Maintenance: Not graded. No commit or release date has been recorded. - 55Trust score 55 of 100, Fair
Extracts PDF text locally, with macOS PDFKit and Vision OCR fallbacks and optional AI retry.
RegistryLicenseQualityMaintenanceLicense: Not graded. No license is recorded for this server.Quality: Not graded. No README or tool list has been captured yet.Maintenance: Not graded. No commit or release date has been recorded. - 55Trust score 55 of 100, Fair
Official PDF.co MCP server: convert, merge, split, OCR, search and fill PDFs through the PDF.co API
RegistryLicenseQualityMaintenanceLicense: Not graded. No license is recorded for this server.Quality: Not graded. No README or tool list has been captured yet.Maintenance: Not graded. No commit or release date has been recorded. - 55Trust score 55 of 100, Fair
Quire PDF
Unknown author
Turn JSON into polished PDFs: invoices, receipts, quotes, certificates, or any data.
RegistryLicenseQualityMaintenanceLicense: Not graded. No license is recorded for this server.Quality: Not graded. No README or tool list has been captured yet.Maintenance: Not graded. No commit or release date has been recorded.
How to install one
- Compare the shortlist. pdf-mcp leads this topic. Open two or three listings and compare their tools, transport, and required API keys.
- Copy the config for your client. Each listing has an install guide, for example Claude Desktop, Claude Code, Cursor, or VS Code. The client guides show where each config file lives.
- Scope credentials, then restart. Use the narrowest token or folder the server needs, restart the client, and check that the new tools appear.
PDF MCP questions
Can PDF MCP servers read scanned documents?
Only servers with OCR can. Look for Tesseract, a document AI service, or vision-model support in the server's description. Without OCR, a scanned PDF comes back as empty pages.
Do PDF servers keep tables intact?
Some do, returning tables as Markdown or CSV. Simple text extractors flatten them into lines, which loses column alignment. If tables matter, test the server on one of your own documents before you commit.
What about Word and PowerPoint files?
Several document servers convert DOCX, PPTX, and XLSX to Markdown alongside PDF. Markitdown and Pandoc based servers are the usual choice for mixed office documents.