How microsoft/markitdown works
MarkItDown accepts local files, streams, and URLs, identifies their format, dispatches to a converter, and returns Markdown. The core includes format-specific extraction plus optional cloud analyzers, web/media conversion, and plugins; the OCR plugin augments Office and PDF conversion with LLM vision. The README documents these capabilities, while the MCP package is visible in the tree and its entry point references the core. Some converter implementations and MCP behavior were not excerpted, so their internal links are intentionally limited.
Main entry points: _markitdown.py Conversion engine · __init__.py Converter registry · _base_converter.py Markdown result · _plugin.py OCR registration
Components 23
Input and dispatch
- Conversion engine
packages/markitdown/src/markitdown/_markitdown.py - Stream metadata
packages/markitdown/src/markitdown/_stream_info.py - Converter contract
packages/markitdown/src/markitdown/_base_converter.py - Converter registry
packages/markitdown/src/markitdown/converters/__init__.py - Markdown result
packages/markitdown/src/markitdown/_base_converter.py
Built-in conversion
- Office formats
packages/markitdown/src/markitdown/converters - PDF conversion
packages/markitdown/src/markitdown/converters/_pdf_converter.py - Web and feeds
packages/markitdown/src/markitdown/converters - Images and audio
packages/markitdown/src/markitdown/converters - Text and archives
packages/markitdown/src/markitdown/converters
Cloud analysis
- Document Intelligence
packages/markitdown/src/markitdown/converters/_doc_intel_converter.py - Content Understanding
packages/markitdown/src/markitdown/converters/_cu_converter.py
OCR plugin
- OCR registration
packages/markitdown-ocr/src/markitdown_ocr/_plugin.py - PDF OCR
packages/markitdown-ocr/src/markitdown_ocr/_pdf_converter_with_ocr.py - DOCX OCR
packages/markitdown-ocr/src/markitdown_ocr/_docx_converter_with_ocr.py - PPTX OCR
packages/markitdown-ocr/src/markitdown_ocr/_pptx_converter_with_ocr.py - XLSX OCR
packages/markitdown-ocr/src/markitdown_ocr/_xlsx_converter_with_ocr.py - LLM vision OCR
packages/markitdown-ocr/src/markitdown_ocr/_ocr_service.py
User interfaces
- Command line
packages/markitdown/src/markitdown/__main__.py - MCP server
packages/markitdown-mcp/src/markitdown_mcp/__main__.py
Other components
- Caller
- Azure services
- Vision API
Connections 31
- Caller → Command line: runs
README.md - Caller → Conversion engine: calls
README.md - Caller → MCP server: uses
- Command line → Conversion engine: invokes
packages/markitdown/src/markitdown/__main__.py - MCP server → Conversion engine: imports
packages/markitdown-mcp/src/markitdown_mcp/__main__.py - Conversion engine → Stream metadata: builds
packages/markitdown/src/markitdown/_markitdown.py - Conversion engine → Converter registry: dispatches
packages/markitdown/src/markitdown/_markitdown.py - Conversion engine → Converter contract: registers
packages/markitdown/src/markitdown/_markitdown.py - Converter registry → Office formats: exports
packages/markitdown/src/markitdown/converters/__init__.py - Converter registry → PDF conversion: exports
packages/markitdown/src/markitdown/converters/__init__.py - Converter registry → Web and feeds: exports
packages/markitdown/src/markitdown/converters/__init__.py - Converter registry → Images and audio: exports
packages/markitdown/src/markitdown/converters/__init__.py - Converter registry → Text and archives: exports
packages/markitdown/src/markitdown/converters/__init__.py - Conversion engine → Document Intelligence: registers
README.md - Conversion engine → Content Understanding: registers
README.md - Document Intelligence → Azure services: requests analysis
README.md - Content Understanding → Azure services: requests analysis
README.md - Conversion engine → OCR registration: loads optionally
README.md - OCR registration → PDF OCR: registers
packages/markitdown-ocr/src/markitdown_ocr/_plugin.py - OCR registration → DOCX OCR: registers
packages/markitdown-ocr/src/markitdown_ocr/_plugin.py - OCR registration → PPTX OCR: registers
packages/markitdown-ocr/src/markitdown_ocr/_plugin.py - OCR registration → XLSX OCR: registers
packages/markitdown-ocr/src/markitdown_ocr/_plugin.py - PDF OCR → LLM vision OCR: extracts text
packages/markitdown-ocr/src/markitdown_ocr/_pdf_converter_with_ocr.py - DOCX OCR → LLM vision OCR: extracts text
packages/markitdown-ocr/src/markitdown_ocr/_docx_converter_with_ocr.py - LLM vision OCR → Vision API: sends image
packages/markitdown-ocr/src/markitdown_ocr/_ocr_service.py - Conversion engine → Markdown result: returns Markdown
README.md - Office formats → Markdown result: produces Markdown
README.md - PDF conversion → Markdown result: produces Markdown
README.md - Web and feeds → Markdown result: produces Markdown
README.md - Images and audio → Markdown result: produces Markdown
README.md - Text and archives → Markdown result: produces Markdown
README.md