# MarkItDown Web Studio

> MarkItDown Web Studio is an executive-grade, private document conversion platform and text processing toolkit for AI pipelines, RAG systems, and knowledge workers. Powered by Microsoft MarkItDown, Apple Vision neural OCR, and FastAPI.

## Core Capabilities
- Universal Document to Markdown: Converts PDF, Word (.docx), Excel (.xlsx, CSV), PowerPoint (.pptx), images (PNG, JPG, WebP), and live web URLs into clean GitHub Flavored Markdown (GFM).
- Apple Vision Neural OCR: Local offline optical character recognition on macOS with zero cloud telemetry.
- AI Markdown Sanitizer: Strips conversational filler, thinking tags, and escape anomalies from ChatGPT, Claude, and Gemini outputs.
- Table Auto-Aligner & TOC Generator: Normalizes ragged pipe tables and generates nested anchor-linked tables of contents.
- Zero Data Retention: All processing runs locally or in-memory; files are wiped immediately upon response.

## Available Web Tools
- [PDF to Markdown](https://markitdown.site/pdf-to-markdown): Extract text, tables, and scanned pages with OCR.
- [Word to Markdown](https://markitdown.site/word-to-markdown): Convert .docx preserving formatting and footnotes.
- [Excel to Markdown](https://markitdown.site/excel-to-markdown): Multi-sheet workbooks into clean Markdown tables.
- [PowerPoint to Markdown](https://markitdown.site/powerpoint-to-markdown): Slide decks into structured notes.
- [Image to Markdown (OCR)](https://markitdown.site/image-to-markdown): Extract text and code snippets from screenshots.
- [URL to Markdown](https://markitdown.site/url-to-markdown): Clean article web clipper stripping ads and navigation.
- [HTML to Markdown](https://markitdown.site/html-to-markdown): Raw HTML into clean semantic GFM.
- [Table Formatter](https://markitdown.site/table-formatter): Auto-align and beautify messy pipe tables.
- [AI Markdown Cleaner](https://markitdown.site/ai-cleaner): Sanitize ChatGPT and Claude outputs.
- [TOC Generator](https://markitdown.site/toc-generator): Clickable hierarchical tables of contents.
- [Markdown to HTML](https://markitdown.site/markdown-to-html): Standalone HTML documents with modern styling.
- [Markdown to PDF](https://markitdown.site/markdown-to-pdf): Print-ready A4 PDF export with executive typography.

## API Endpoints
- `POST https://markitdown.site/api/convert/file`: Multipart upload (`file`). Accepts PDF, DOCX, XLSX, PPTX, images.
- `POST https://markitdown.site/api/convert/url`: JSON `{"url": "https://example.com"}`.
- `POST https://markitdown.site/api/convert/html`: JSON `{"html": "<p>Content</p>"}`.
- `POST https://markitdown.site/api/process/table`: JSON `{"markdown": "|col|..."}`.
- `POST https://markitdown.site/api/process/clean`: JSON `{"markdown": "AI output..."}`.
- `POST https://markitdown.site/api/process/toc`: JSON `{"markdown": "# Title..."}`.

## Full Specification
- [Full LLM Specification](https://markitdown.site/llms-full.txt)