Mdkit
Converts PDFs, Office documents, and HTML into clean, LLM-ready Markdown and JSON for AI applications.
Mdkit transforms various document types, including PDFs, Word, PowerPoint, Excel, and HTML, into structured, clean Markdown and JSON. This tool is designed to provide 'LLM-ready' output, making it ideal for directly feeding real-world documents into large language models or RAG (Retrieval Augmented Generation) pipelines without requiring extensive manual data cleaning. It offers access via a simple REST API and an MCP server endpoint, supporting both synchronous conversions for smaller files and an asynchronous job queue with polling and webhooks for larger tasks. Mdkit positions itself as a robust solution for developers needing to integrate document understanding into their AI workflows.