NetMind ParsePro is an MCP server that parses PDF documents from local storage paths or remote web URLs into structured formats like JSON and Markdown. It connects AI agent environments directly to NetMind's cloud-based PDF parsing service using the Model Context Protocol. Software developers, data analysts, and researchers use this server to read and extract information from PDF files without requiring custom local optical character recognition or manual text extraction scripts. By handling both local file systems and remote HTTP or HTTPS links, the tool allows AI agents like Claude or Cursor to inspect technical documentation, academic papers, and enterprise reports. Extracted data can be returned either as a structured JSON object or as cleanly formatted Markdown strings, ensuring LLMs can interpret and reason over complex document structures, tables, and text sections efficiently during automated workflows.
Category: Developer Tools & Code Intelligence
Tags: extraction, json, markdown, parsing, pdf
uv installed on your machine using your system's package manager or installation script. 2. Obtain an API key from your NetMind account dashboard. 3. Open your client's MCP configuration file (such as claude_desktop_config.json for Claude Desktop or mcp.json for Cursor and Windsurf). 4. Add the parse-pdf server definition using uvx: json { "mcpServers": { "parse-pdf": { "env": { "NETMIND_API_TOKEN": "YOUR_NETMIND_API_KEY" }, "command": "uvx", "args": [ "netmind-parse-pdf-mcp" ] } } } 5. Save the configuration file and restart your MCP client.Part of MCP Servers
You can run NetMind ParsePro directly using the uvx runner. Install uv on your operating system, obtain a NetMind API token, and add the server definition to your MCP client configuration file under the parse-pdf server entry.
The server provides the parse_pdf tool, which takes a local file path or remote URL pointing to a PDF file. It parses the document content and returns it formatted either as structured JSON or as a clean Markdown string.
NetMind ParsePro works with any standard Model Context Protocol client, including Claude Desktop, Cursor, and Windsurf, by adding the uvx execution configuration to the respective client's JSON configuration file.
The tool accepts a source parameter, which can be an absolute local file path or a remote URL starting with http:// or https://. It also accepts a format parameter set to either json or markdown.