playground

What it does

0xpdfmcp is a Model Context Protocol (MCP) server that enables AI agents to intelligently parse PDF documents into structured JSON format. It uses a schema-first approach where you define the JSON structure you want to extract, then the tool automatically parses invoices, forms, and other documents to match that schema. The service includes optional OCR capabilities for scanned documents and supports asynchronous job polling for large-scale document processing.

How to set up

Installation is straightforward using uvx: run uvx oxpdfmcp to launch the MCP server. The tool works across multiple platforms (Python, Cloud, Mac, Windows, Linux) and integrates with the 0xPdf API. Complete setup documentation is available in the official docs at 0xpdf.io/docsmcp, with a playground environment at 0xpdf.io/samples for testing before integration.

Key features

• Schema-first PDF parsing with JSON output validation
• Support for invoices, forms, and complex document layouts
• Optional OCR for scanned and image-based PDFs
• Schema management and versioning capabilities
• Async job polling for batch document processing
• Multi-platform compatibility and cloud deployment support