MergePDF
Merge PDF View All Tools →
Model Context Protocol 1.0 Ready

MergePDF.pro MCP Server

Empower your AI assistants (Claude Desktop, Cursor, Windsurf, agentic workflows) with native PDF manipulation, extraction, and document conversion tools.

1-Click NPX CLI

Zero Setup

Run directly via Node.js without cloning or installing dependencies:

npx -y @mergepdf/mcp-server

Lightweight stdio client connecting any local agent to MergePDF tools.

Claude Desktop

claude_desktop_config.json

Add the MergePDF.pro HTTP SSE endpoint to your Claude Desktop configuration:

{
  "mcpServers": {
    "mergepdf": {
      "url": "https://mergepdf.pro/mcp/tools"
    }
  }
}

Claude can now inspect, merge, compress, and extract text from PDFs directly in chat.

Cursor / Windsurf

.cursor/mcp.json

Register the MergePDF tools in your project or global Cursor settings:

{
  "mcpServers": {
    "mergepdf": {
      "url": "https://mergepdf.pro/mcp/tools"
    }
  }
}

Allows AI coding agents to convert documents and inspect files during build tasks.

Smithery.ai CLI

Package Registry

Install directly via the Smithery CLI into Claude Desktop or Zed:

npx -y @smithery/cli install mergepdf --client claude

Configured with smithery.yaml and discoverable in global MCP registries.

Developer API Access & Quotas

Need 500 req/min for Autonomous Agents?

Anonymous calls are limited to 60 req/min. Generate a free developer API key to unlock 500 req/min for high-throughput batch operations and autonomous agents.

Live Test Console

Interactive MCP Tool Tester

Simulate an AI agent dispatching a tools/call JSON-RPC 2.0 request against our live MCP endpoint.

Ready
Endpoint: POST https://mergepdf.pro/mcp/tools
Protected by internal SSRF blocking against private/loopback IP addresses and throttled per IP.
Terminal Response
// Select a tool and click "Send JSON-RPC Request" to inspect live response...
// Response format follows official Model Context Protocol 1.0 JSON-RPC specification.

Available Agent Tools

Every tool returns structured data and verified download links.

list_pdf_tools

Discovers all 60+ online PDF and image tools available on MergePDF.pro with URLs.

Arguments: none
get_pdf_info

Extracts metadata: page count, title, author, creation date, dimensions, and encryption state.

Arguments: url (string)
extract_pdf_text

Extracts clean plain text from a remote PDF directly into the LLM context for reasoning.

Arguments: url (string)
merge_pdfs

Merges two or more PDF documents into a single file and returns a download link.

Arguments: urls (array of strings)
compress_pdf

Compresses bloated documents with customizable levels (recommended, extreme, low).

Arguments: url, compression_level
convert_pdf_to_word

Converts PDF into an editable Microsoft Word document (.docx) with download link.

Arguments: url (string)
convert_pdf_to_excel

Extracts tabular data from statements or reports into Microsoft Excel (.xlsx).

Arguments: url (string)
pdf_to_markdown

Converts PDF into structured Markdown with headings and tables directly for LLM context.

Arguments: url (string)
ocr_pdf

Runs Tesseract OCR on scanned PDFs, generating a searchable PDF and recognized text excerpt.

Arguments: url (string), language (string, optional)
split_pdf

Extracts specific pages or page ranges (e.g. 1-5, 8) into a separate PDF document.

Arguments: url (string), pages (array or string)
rotate_pdf

Rotates all or selected document pages clockwise by 90, 180, or 270 degrees.

Arguments: url (string), angle (integer), pages (array, optional)
watermark_pdf

Applies custom text stamps (CONFIDENTIAL, DRAFT, copyright) with controllable opacity.

Arguments: url (string), text (string), opacity, rotation
protect_pdf

Encrypts documents with industry-standard AES-256 encryption and opening password.

Arguments: url (string), password (string)
unlock_pdf

Removes password protection and owner permissions restrictions from encrypted PDFs.

Arguments: url (string), password (string, optional)

Security & Privacy by Design

MergePDF.pro enforces strict Server-Side Request Forgery (SSRF) filters. All remote URLs are validated to prevent connections to private, loopback, or internal networks.

Processed files are isolated and automatically deleted according to our data privacy policy. Files are never used for AI model training.