MarkitaiClean Markdown from files and URLs

Convert documents, images, and web pages with one command. LLM enhancement stays optional.

Multi-format Support
Convert DOCX, PPTX, XLSX, PDF, EPUB, EML, TXT, MD, images (JPG/PNG/WebP) and URLs to Markdown. Legacy .doc/.ppt via the legacy extra.
LLM Enhancement
AI-powered format cleaning, metadata generation (frontmatter), and image analysis.
Batch Processing
Concurrent conversion with progress display and resume capability for interrupted jobs.
OCR Recognition
Extract text from scanned PDFs and images with local RapidOCR (markitai[ocr]), or let a vision model read the pages directly with --ocr --llm.
MCP for AI Agents
A bundled MCP server exposes convert_document, convert_url, batch_convert and job_status over stdio, so agents convert files without loading markitai into their context.
Output Profiles
Shape the result for its consumer with --profile rag, obsidian or okf: visible assets, wikilinks or OKF frontmatter, without changing the conversion itself.