YouTube Transcript

Youtube Transcript MCP Server
  • python

1

GitHub Stars

python

Language

6 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs
{
  "mcpServers": {
    "kyungpyo-kim-youtube_transcript_mcp": {
      "command": "/path/to/your/youtube_transcript_mcp/venv/bin/python",
      "args": [
        "/path/to/your/youtube_transcript_mcp/youtube_transcript_mcp_server.py"
      ]
    }
  }
}

You can run a dedicated MCP server to fetch YouTube video transcripts, including multilingual and auto-generated captions, for use with AI models like Claude. This server exposes simple tools to retrieve transcripts and to list available transcripts for a given YouTube URL, enabling seamless integration with your workflows.

How to use

After you have the server running, you can interact with it through an MCP client to perform two main actions. First, get a transcript for a YouTube video in your preferred language. Second, list the available transcripts for that video so you know which languages or caption types are accessible.

If you want the Korean default transcript, you can request the transcript for a video URL and specify the language code when supported. You can also request English specifically or query the system for all available transcripts to decide which language to use.

How to install

Prerequisites you need before installation are Python and a compatible environment to run a Python-based MCP server.

Step 1: Clone the project and navigate into the directory.

git clone <repository-url>
cd youtube_transcript_mcp

Step 2: Create and activate a virtual environment.

python3 -m venv venv
source venv/bin/activate  # Linux/Mac
# 또는 Windows의 경우: venv\Scripts\activate

Step 3: Install dependencies.

pip install -r requirements.txt

Additional configuration and notes

Configure Claude Desktop to use this MCP server by pointing to the Python executable inside your virtual environment and the server script. You will add an entry under your MCP servers with the appropriate paths to run the server locally.

{
  "mcpServers": {
    "youtube-transcript": {
      "command": "/path/to/your/youtube_transcript_mcp/venv/bin/python",
      "args": ["/path/to/your/youtube_transcript_mcp/youtube_transcript_mcp_server.py"],
      "env": {}
    }
  }
}

Important: Replace "/path/to/your/youtube_transcript_mcp" with the actual filesystem path where you cloned the project.

Troubleshooting and tips

If changes do not take effect, restart Claude Desktop after updating the configuration. Ensure the configuration file path is correct and that the virtual environment is activated when starting the server.

If a transcript cannot be found, the video may lack captions, be private, or have access restrictions. You can use the list transcripts tool to verify which languages are available for that video.

Supported URLs and languages

Supported YouTube URL formats include the standard watch URL, shortened URLs, or a 11-character video ID.

Default language is Korean (ko). English (en) is also supported, along with any other language YouTube provides.

If a language is not explicitly requested, the server will attempt Korean → English → auto-generated captions in order.

API tools

This MCP server exposes two tools for clients to call.

get_youtube_transcript — Retrieve the transcript for a YouTube video. Parameters: url (required), language (optional, default: ko).

list_available_transcripts — Retrieve the list of transcripts available for a YouTube video. Parameters: url (required).

Available tools

get_youtube_transcript

Fetches the transcript for a given YouTube video URL in the specified language, defaulting to Korean.

list_available_transcripts

Returns the list of transcripts available for a given YouTube video URL, helping you see which languages and caption types are accessible.

Built by
VeilStrat
AI signals for GTM teams
© 2026 VeilStrat. All rights reserved.All systems operational