PDF Reader MCP

Overview

PDF Reader MCP is a Model Context Protocol server that brings evidence-first PDF analysis to any MCP-compatible AI assistant. It acts as an Agent Document Twin, extracting and citing specific pages and bounding boxes from PDFs so your AI can reference exact locations in documents. Any MCP client can call its tools directly during conversations and agent runs.

The server runs locally via the npm package @sylphx/pdf-reader-mcp, installed with npx -y @sylphx/pdf-reader-mcp. It requires no authentication and works immediately after installation. The latest release is v4.1.2, published under the MIT license. Source code and documentation are available at SylphxAI/pdf-reader-mcp on GitHub, where the project has 907 stars.

PDF Reader MCP is categorized as AI & Agents on mcp.site and integrates with Claude Desktop, Claude Code, Cursor, VS Code, Windsurf, Zed, and other MCP clients. If you maintain this server, claim your listing to earn the Verified badge and keep information up to date.

Installation

// claude_desktop_config.json
{
  "mcpServers": {
    "pdf_reader_mcp": {
      "command": "npx",
      "args": [
        "-y",
        "@sylphx/pdf-reader-mcp"
      ]
    }
  }
}