YouTube Transcript MCP Server
A high-performance, serverless implementation of a YouTube transcript extraction service using the Model Context Protocol (MCP), deployed on Cloudflare Workers.
Overview
This MCP server enables AI assistants to retrieve transcripts from YouTube videos through a simple API. The implementation combines the lightweight transcript extraction capabilities seen in kimtaeyoon83/mcp-server-youtube-transcript with the remote MCP server architecture from Cloudflare AI demos.
Features
- Serverless Deployment: Runs on Cloudflare's global edge network for minimal latency
- YouTube URL Flexibility: Supports multiple URL formats and direct video IDs
- Language Selection: Retrieve transcripts in different languages (defaults to English)
- Edge-optimized: Ultra-fast response times (typically 400-800ms)
- Minimal Implementation: Less than 300 lines of code for easy maintenance
- SSE Transport: Implements Server-Sent Events for streaming connections
Available Tools
| Tool | Description | Parameters |
|---|---|---|
get_transcript |
Extract transcripts from YouTube videos | url (required): YouTube video URL or IDlang (optional, default: "en"): Language code |
Usage with Claude Desktop
Installation
- Open Claude Desktop and go to Settings > Developer > Edit Config
- Update your configuration file:
{
"mcpServers": {
"youtube-transcript": {
"command": "npx",
"args": [
"mcp-remote",
"https://your-deployed-worker.workers.dev/sse"
]
}
}
}
- Restart Claude Desktop
Example Prompts
Can you show me the transcript of this YouTube video? https://www.youtube.com/watch?v=dQw4w9WgXcQ
Extract the transcripts from this TED talk and summarize the key points: https://youtu.be/8S0FDjFBj8o
Local Development
Prerequisites
- Node.js 18 or higher
- Wrangler CLI (
npm install -g wrangler)
Setup
# Clone the repository
git clone https://github.com/yourusername/youtube-transcript-mcp-server.git
cd youtube-transcript-mcp-server
# Install dependencies
npm install
# Run locally
wrangler dev
Testing with MCP Inspector
-
Install the MCP Inspector:
npx @modelcontextprotocol/inspector -
In the inspector:
- Set Transport Type to
SSE - Enter
http://localhost:8787/sseas the URL - Click "Connect"
- Set Transport Type to
-
Try out the
get_transcripttool with different YouTube URLs
Deployment to Cloudflare
# Deploy to Cloudflare Workers
wrangler deploy
Technical Implementation
The server is built with a minimal, high-efficiency codebase:
- YouTubeTranscriptMCPSqlite: Core MCP agent implementation with transcript extraction capabilities
- McpServer: Handles MCP protocol interactions
- MCP Protocol Integration: Complete implementation of the Model Context Protocol
Key optimizations:
- Efficient URL parsing and validation
- Minimal external dependencies
- Proper error handling with detailed error messages
- Streaming support through SSE
Credits
- Built with @modelcontextprotocol/sdk
- Uses youtube-captions-scraper for transcript extraction
- Inspired by:
License
MIT
Recommend MCP Servers 💡
cbi-mcp-server
Provides an interface for developers to interact with CB Insights ChatCBI LLM through AI Agents.
cartesia-mcp
The Cartesia MCP server enables clients like Cursor, Claude Desktop, and OpenAI agents to interact with Cartesia's API for speech localization, text-to-audio conversion, and voice infill.
lingo.dev
A MCP server for Lingo.dev that enables AI tools to translate apps, websites, and data using LLMs via the Model Context Protocol
anki-mcp-server
An MCP server that connects to a locally running Anki desktop app via the Anki-Connect add-on, providing functionalities for card review, creation, and search.
xinlei413/DOC-Server-MCP
An MCP server designed to scrape, process, index, and search documentation for various software libraries and packages, supporting multiple LLM embedding providers and hybrid search.
devrev-mcp
A Model Context Protocol server for DevRev, providing comprehensive access to DevRev's APIs for managing work items, parts, meetings, workflows, and search across DevRev data.