MCP Video Digest

🚀 MCP Video Transcription Tool

The MCP Video Transcription Tool is a powerful video processing software that supports downloading videos from YouTube and other platforms and converting their audio into text. This tool integrates a variety of advanced speech recognition technologies, enabling users to quickly complete video transcription tasks.

✨ Features

• Multi-platform Compatibility: Supports mainstream video platforms such as YouTube and Vimeo.
• Efficient Download: Equipped with a powerful built-in downloader capable of handling various video formats.
• Accurate Transcription: Integrates well-known speech recognition services like Deepgram, Gladia, AssemblyAI, and Speechmatics.
• Flexible Configuration: Allows for custom API keys and parameter settings.
• Intelligent Error Handling: Automatically detects and fixes issues that occur during the download and transcription processes.

📦 Project Structure

MCP-Video/
├── src/
│   ├── services/
│   │   ├── download/          # Download module
│   │   │   └── __init__.py
│   │   └── transcription/     # Transcription module
│   │       └── __init__.py
├── tests/
│   ├── test_env.py           # Environment test
│   ├── test_downloader.py    # Downloader test
│   └── test_transcription.py  # Transcriber test
└── requirements.txt          # Dependency management file

📦 Installation

Installation Steps

1. Clone the project repository:

git clone https://github.com/your-username/MCP-Video.git
cd MCP-Video

2. Install the dependencies:

uv pip install -r requirements.txt

📚 Configuration

Environment Variable Configuration

Create a config.env file in the project root directory and fill in the following content:

DEEPGRAM_API_KEY=your_key_here
GLADIA_API_KEY=your_key_here
ASSEMBLYAI_API_KEY=your_key_here
SPEECHMATICS_API_KEY=your_key_here

Service Priority Setting

# config.py
SERVICE_PRIORITY = [
    'Deepgram',
    'Gladia',
    'AssemblyAI',
    'Speechmatics'
]

💻 Usage Examples

Basic Usage

uv run main.py

Advanced Usage

from mcp_video import MCPVideoProcessor

processor = MCPVideoProcessor()
# Download the video
video_path = processor.download_video('https://www.youtube.com/watch?v=12345')
# Extract the audio
audio_path = processor.extract_audio(video_path)
# Transcribe the audio
transcript = processor.transcribe(audio_path)
print(transcript.text)

📚 Testing

Run Tests

uv test

📚 Development Guide

Adding a New Service

Create a new class in the src/services/transcription/ directory, inherit from BaseTranscriptionService, and implement the transcribe method.

Developing a Custom Downloader

Extend or modify the existing downloader in the src/services/download/ directory. It is recommended to refer to the YouTubeDownloader class.

📚 Dependency Management

• Use uv pip install package_name to install new dependencies.
• Use uv pip freeze > requirements.txt to export the current dependencies.

🔧 Error Handling

The service has a well-established error handling mechanism that can automatically handle the following issues:

• Failed API requests
• Interrupted network connections
• Exceeded resource quotas
• File operation exceptions

📚 Notes

⚠️ Important Note

• Ensure that there is sufficient disk space on the system to support temporary file storage.
• It is recommended to use Python 3.11 or a higher version.
• Additional authentication configuration may be required for YouTube downloads.
• Manage API call counts reasonably to avoid triggering rate limits.

📚 Free Quota Information

• Speechmatics: 8 hours of free usage per month.
• Gladia: 10 hours of free usage per month.
• AssemblyAI: A free quota of $50 per month.
• Deepgram: A free quota of $200 per month.

📄 License

This project follows the MIT open-source license. For details, please refer to the LICENSE file.