markitdown-ui
A local web interface and MCP server for converting documents (PDF, Word, Excel, images, audio, etc.) to Markdown using Microsoft's MarkItDown library.
README
⚡ Cyberpunk MarkItDown Document Converter & GUI
A high-contrast, retro-futuristic local web interface for converting complex documents (PDFs, Word documents, Excel sheets, presentations, audio recordings, images, and HTML) into clean, structured Markdown.
This local tool provides a graphical user interface (GUI) wrapper on top of Microsoft's official MarkItDown Python library.
📸 Interface Preview

🔍 Core Technology Disclosure
This project is a wrapper interface and does not compile documents from scratch. It relies directly on:
- Microsoft MarkItDown: The core document conversion engine. Visit their repository at github.com/microsoft/markitdown for more details.
- Model Context Protocol (MCP) Compatible: The same core parser running this interface is also configured as an MCP server for AI clients (like Claude Desktop).
🛠 Features
- Visual Design: Immersive neon cyber-terminal styled with retro CRT scanline grids, glitch logo, and glowing interactive elements.
- Interactive Audio: Real-time feedback sounds (beeps, success chimes, error hums) generated on-the-fly using the Web Audio API (no external audio assets required).
- Live Diagnostic Feed: Real-time log console showing exactly what files are loaded, processed, and deleted.
- Multi-Format Extraction: Seamlessly translates text and structural data from PDF, DOCX, XLSX, PPTX, HTML, CSV, PNG/JPG, and Audio (Speech-to-Text).
⚙️ Step-by-Step Installation Guide
To get this app running on your computer, follow these simple steps:
⚙️ Installation & Setup
Choose your operating system to set up and run the decompiler:
🍎 macOS Installation Guide
1. Install System Dependencies
We use Homebrew to manage command-line installations cleanly. Open your terminal and run:
# Install Homebrew (if you don't have it)
/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)"
# Install Python 3, pipx, and Node.js
brew install python pipx node
# Configure pipx path (requires terminal restart after running)
pipx ensurepath
2. Install Microsoft MarkItDown
Install the conversion engine globally with all optional feature formats (PDF, Excel, images, audio, etc.):
pipx install "markitdown[all]"
Verify the installation works by running: markitdown --help
3. Clone and Start the Server
# Clone the repository
git clone https://github.com/shahariarovi789-ux/markitdown-ui.git
cd markitdown-ui
# Install dependencies
npm install
# Start the server
npm start
Open http://localhost:3000 in your browser.
🪟 Windows Installation Guide
1. Install System Dependencies
We recommend using Windows Package Manager (winget) or installing manually. Open PowerShell and run:
# Install Python, Node.js, and Git
winget install -e --id Python.Python.3
winget install -e --id OpenJS.NodeJS
winget install -e --id Git.Git
Note: Make sure to restart your terminal or computer after installing these packages so the environment variables take effect.
2. Configure pipx and Install Microsoft MarkItDown
Open a new PowerShell window and run:
# Install pipx via Python pip
py -m pip install --user pipx
# Configure pipx paths
py -m pipx ensurepath
Note: Please restart your PowerShell window after running ensurepath.
# Install Microsoft MarkItDown with all features
pipx install "markitdown[all]"
Verify the installation works by running: markitdown --help
3. Clone and Start the Server
# Clone the repository
git clone https://github.com/shahariarovi789-ux/markitdown-ui.git
cd markitdown-ui
# Install dependencies
npm install
# Start the server
npm start
Open http://localhost:3000 in your browser.
🐧 Linux Installation Guide (Ubuntu / Debian / Fedora / Arch)
1. Install System Dependencies
Install Python 3, pipx, Node.js, and Git using your distribution's package manager:
Ubuntu / Debian:
sudo apt update
sudo apt install -y python3 python3-pip python3-venv pipx nodejs npm git
Fedora:
sudo dnf install -y python3 python3-pip pipx nodejs npm git
Arch Linux:
sudo pacman -S python python-pip pipx nodejs npm git
Configure pipx path (requires terminal restart after running):
pipx ensurepath
2. Install Microsoft MarkItDown
Install the conversion engine globally with all optional feature formats (PDF, Excel, images, audio, etc.):
pipx install "markitdown[all]"
Verify the installation works by running: markitdown --help
3. Clone and Start the Server
# Clone the repository
git clone https://github.com/shahariarovi789-ux/markitdown-ui.git
cd markitdown-ui
# Install dependencies
npm install
# Start the server
npm start
Open http://localhost:3000 in your browser.
📁 Repository Structure
markitdown-ui/
├── public/
│ └── index.html # Self-contained Cyberpunk UI frontend
├── uploads/ # Temporary directory for file conversion buffers (auto-cleaned)
├── server.js # Express API backend server
├── screenshot.png # Application screenshot
├── package.json # Dependencies & start scripts
├── .gitignore # Git exclusion rules
└── README.md # Documentation
🔧 Troubleshooting
1. markitdown CLI command not found
If your terminal states that the command markitdown cannot be found, it means ~/.local/bin (where pipx installs command symlinks) is not in your system $PATH.
- Run
pipx ensurepathagain. - If that doesn't fix it, manually add it to your shell configuration file (
~/.bashrc,~/.zshrc, or~/.bash_profile):export PATH="$HOME/.local/bin:$PATH" - Open a new terminal and try running
markitdown --help.
2. Conversions fail or return empty output
Ensure that you ran pipx install "markitdown[all]" instead of just pipx install markitdown. The [all] tag installs essential parsing libraries like PyMuPDF for PDFs, openpyxl for Excel files, and others required to successfully unpack formats.
- To upgrade your existing installation, run:
pipx install --force "markitdown[all]"
💻 Running in the Background
To run the web interface in the background without keeping a terminal open:
nohup npm start > server.log 2>&1 &
To stop the background server:
kill $(lsof -t -i:3000)
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。