MCP YouTube Extract

MCP YouTube Extract is an MCP server that retrieves video metadata and subtitles directly from YouTube without requiring an API key. It connects MCP-compatible clients to YouTube using the yt-info-extract and yt-ts-extract libraries, providing built-in fallback routines to handle both manual and auto-generated transcripts. Software developers, content analysts, and researchers use this tool to feed structured YouTube data into language models for automated summarization, study, and content creation. The server exposes the get_yt_video_info tool, which accepts a YouTube video ID and returns critical metadata including title, channel details, publish date, view count, full description, and complete transcripts. While basic functionality runs out of the box through web extraction techniques, users can optionally configure a YouTube Data API v3 key to serve as an additional metadata fallback. Through its error-handling and logging architecture, MCP YouTube Extract streamlines the ingestion of online video materials into automated AI agent workflows.

Category: Design, Media & Creative

Tags: media, metadata, transcription, video, youtube

Visit MCP YouTube Extract

How to install and configure MCP YouTube Extract

1. Install the package using pip or pipx: bash pip install mcp-youtube-extract 2. Open your client configuration file (such as Claude Desktop or Cursor). 3. Add the server entry to the mcpServers object: json { "mcpServers": { "mcp_youtube_extract": { "command": "mcp_youtube_extract" } } } 4. Optional: To supply a YouTube Data API key as an extra fallback, add the environment variable inside the configuration: json { "mcpServers": { "mcp_youtube_extract": { "command": "mcp_youtube_extract", "env": { "YOUTUBE_API_KEY": "your_youtube_api_key" } } } } 5. Restart your client application to load the server.

What you can do with MCP YouTube Extract

• Summarizing long YouTube lectures and podcasts by extracting auto-generated or manual video transcripts into an AI chat context. - Automating research workflows by collecting video publish dates, descriptions, channel names, and view counts via video IDs. - Creating automated study notes and study guides based on educational YouTube presentations without manual transcript copying. - Extracting timestamped dialogue from interviews and panels to assist content creators with clip generation and quote referencing.

Key facts

• https://github.com/sinjab/mcp_youtube_extract
• Design, Media & Creative, Web Search & Research
• media, metadata, transcription, video, youtube

Part of MCP Servers

How do I install MCP YouTube Extract?

You can install MCP YouTube Extract directly from PyPI using pip install mcp-youtube-extract or pipx install mcp-youtube-extract. After installing, add the mcp_youtube_extract command to your client MCP configuration file, then restart the client to access its tools.

Do I need a YouTube API key to use MCP YouTube Extract?

No, an API key is not required for basic functionality. The server works out of the box using yt-info-extract and yt-ts-extract. However, you can optionally configure a YouTube Data API v3 key in your environment settings as a fallback for metadata queries.

What tools does MCP YouTube Extract provide?

The server provides the get_yt_video_info tool. By passing a YouTube video ID, it returns metadata like video title, channel name, publication date, view count, and description, alongside video transcripts with fallback support for manual and auto-generated captions.

What license does MCP YouTube Extract use?

MCP YouTube Extract is open-source software licensed under the MIT License. You can inspect the source code, fork the project, and contribute directly via its GitHub repository.