Youtube Vision

🚀 YouTube Video MCP Service (youtube-vision)

YouTube Vision is a video analysis tool based on Google Gemini Vision API, capable of extracting and understanding visual information from video content. This service supports multiple functions, including generating descriptions, summarizing video content, answering questions, and identifying key segments.

🚀 Quick Start

YouTube Vision is a powerful video analysis tool. You can quickly start using it through the following installation methods.

✨ Features

• Video Analysis: Analyze YouTube videos using the Gemini Vision API.
• Description Generation: Automatically generate text descriptions of videos.
• Content Summarization: Extract the core information of videos and generate summaries.
• Q&A Function: Answer specific questions based on video content.
• Key Segment Identification: Identify and describe important parts of videos.

📦 Installation

1. Install via Smithery

Smithery is a platform for deploying and managing MaaS (Model as a Service), which supports the rapid deployment of the YouTube Vision service. Follow these steps:

• Register and log in to the Smithery platform.
• Search for "YouTube Vision" in the service marketplace.
• Complete the creation and configuration of the service as prompted.

2. Quickly Run with npx

If you have Node.js installed, you can use the following command to quickly start the service:

npx youtube-vision

3. Manually Install from Source Code

Install Dependencies

Use npm to install the required dependencies:

npm install youtube-vision

Start the Service

Run the main file to start the service:

node index.js

📚 Documentation

Configuration Instructions

Environment Variables

• GOOGLE_API_KEY: The key for accessing Google APIs, which needs to be obtained from Google Cloud Console.
• PORT: Specify the port on which the service runs, with a default of 3000.

Functional Modules

1. Video Description Generation

Input

• video_url (required): YouTube video URL.
• language (optional, string): Output language, default is English.

Output

• A text description of the video.

2. Content Summarization

Input

• video_url (required): YouTube video URL.
• summary_length (optional, string): Summary length options ('short', 'medium', 'long'), default is 'medium'.

Output

• A text summary of the video.

3. Key Segment Identification

Input

• video_url (required): YouTube video URL.
• num_clips (optional, integer): The number of key segments to extract, default is 3.

Output

• A list of key segments containing timestamps and descriptions.

4. Frequently Asked Questions

Input

• video_url (required): YouTube video URL.
• question (required): A specific question.

Output

• A text answer, which may include links to reference materials.

Notes

⚠️ Important Note

• Model Selection: Choose the appropriate analysis module based on the video content and requirements.
• API Limitations: Google APIs have certain limits on the number of calls for the free tier. Please control your usage frequency.
• Content Review: Ensure compliance with relevant laws and regulations before processing sensitive or restricted content.

📄 License

This project is licensed under the MIT License. For more detailed information, please refer to the LICENSE file.