YouTube Search MCP Server
Enables YouTube video search and audio download via web scraping, requiring no API key.
README
YouTube Search API
Zero-cost YouTube video search service powered by web scraping
Features
- ✅ No YouTube API key required - completely free
- ✅ RESTful API design with Swagger documentation
- ✅ MCP (Model Context Protocol) support - integrate with AI assistants
- ✅ Redis caching optimization (1 hour TTL)
- ✅ Complete video metadata extraction
- ✅ Sorting and filtering capabilities
- ✅ Docker containerization support
Quick Start
Install Dependencies
uv sync
Start the Service
python main.py
The service will start at http://localhost:8000.
API Documentation
Visit http://localhost:8000/docs for interactive API documentation.
API Usage Examples
Basic Search
curl "http://localhost:8000/api/v1/search?keyword=Python tutorial"
Specify Result Count
curl "http://localhost:8000/api/v1/search?keyword=Python&limit=5"
Sort by Date
curl "http://localhost:8000/api/v1/search?keyword=Python&sort_by=date&limit=3"
Audio Download Feature
Download YouTube videos as MP3 audio files with automatic caching and cleanup.
Download Features
- ✅ Download YouTube videos as 128kbps MP3 files
- ✅ Support for single and batch downloads
- ✅ Redis-based caching to avoid duplicate downloads
- ✅ Automatic cleanup of expired files (24-hour TTL)
- ✅ Rate limiting: 20 downloads per IP per hour
- ✅ Video duration limit: maximum 10 minutes
- ✅ Direct streaming or download link format
Download API Examples
Single Video Download (Link Format)
# Get download link
curl -X POST "http://localhost:8000/api/v1/download/audio?video_id=dQw4w9WgXcQ&format=link"
# Response
{
"video_id": "dQw4w9WgXcQ",
"title": "Rick Astley - Never Gonna Give You Up",
"duration": 212,
"download_url": "http://localhost:8000/downloads/dQw4w9WgXcQ_Rick_Astley.mp3",
"cached": false,
"file_size": 3400000
}
Single Video Download (Stream Format)
# Direct MP3 stream - returns binary audio data
curl -X POST "http://localhost:8000/api/v1/download/audio?video_id=dQw4w9WgXcQ&format=stream" -o output.mp3
Batch Download
# Download multiple videos at once (max 20)
curl -X POST "http://localhost:8000/api/v1/download/batch" \
-H "Content-Type: application/json" \
-d '{
"video_ids": ["dQw4w9WgXcQ", "jNQXAC9IVRw"]
}'
# Response
{
"total": 2,
"successful": 1,
"failed": 1,
"items": [
{
"video_id": "dQw4w9WgXcQ",
"status": "success",
"download_url": "http://localhost:8000/downloads/...",
"duration": 212,
"cached": false
},
{
"video_id": "jNQXAC9IVRw",
"status": "failed",
"error_message": "Video duration exceeds limit"
}
]
}
Configuration
Set these environment variables to customize download behavior:
# Download storage directory
DOWNLOAD_DIR=/tmp/youtube_audio
# Base URL for serving downloads (used in download links)
DOWNLOAD_BASE_URL=http://localhost:8000/downloads
# Download timeout in seconds
DOWNLOAD_TIMEOUT=300
# Maximum video duration in seconds (default: 600 = 10 minutes)
MAX_VIDEO_DURATION=600
# Audio bitrate in kbps
AUDIO_BITRATE=128
# Cache TTL in hours
CACHE_TTL_HOURS=24
# Rate limit: downloads per IP per hour
RATE_LIMIT_DOWNLOAD_PER_HOUR=20
# Enable rate limiting
RATE_LIMIT_ENABLED=true
Error Handling
The API returns appropriate HTTP status codes:
400- Invalid video ID or parameters403- Video too long, live stream, or access restricted404- Video not found or deleted503- Download failed or YouTube unavailable507- Server storage full
Cleanup
Expired audio files are automatically deleted after 24 hours. Manual cleanup can be triggered:
# Manual cleanup script
python scripts/cleanup_cron.py
# Schedule with cron (daily at 2 AM)
0 2 * * * /usr/bin/python /path/to/scripts/cleanup_cron.py >> /path/to/logs/cleanup.log 2>&1
MCP Integration (Model Context Protocol)
This service provides MCP server functionality, allowing AI assistants (such as Claude Desktop) to directly invoke YouTube search tools via the MCP protocol.
MCP Service Types
- HTTP Service: Provided via StreamableHTTPSessionManager, integrated with existing FastAPI application
- Supported Transport Modes: HTTP (MVP)
- Future Iterations: Consider supporting stdio and SSE modes
REST API and MCP Coexistence
- REST API Preservation: Existing REST API endpoints (
/api/v1/search, etc.) remain fully functional - Flexible Deployment: Choose between integrated (same FastAPI app) or separate (independent process) deployment
- Backward Compatibility: MCP additions do not modify any existing REST API signatures or behaviors
MCP Configuration
Claude Desktop Setup
Add the following to Claude Desktop's configuration file (~/.config/claude/settings.json or %APPDATA%\Claude\settings.json):
{
"servers": {
"youtube-search-mcp": {
"command": "uv",
"args": ["run", "python", "mcp_stdio.py"],
"env": {
"MCP_SEARCH_TIMEOUT": "15",
"MCP_SEARCH_RETRIES": "3",
"PYTHONUNBUFFERED": "1"
}
}
}
}
Environment Variables
MCP_SEARCH_TIMEOUT(default: 15 seconds): Search operation timeoutMCP_SEARCH_RETRIES(default: 3 times): Retry attempts on search failureREDIS_HOST,REDIS_PORT,REDIS_DB: Redis cache configuration (optional)
Starting the MCP Server
# Start MCP server (stdio mode)
python mcp_stdio.py
# Or using uv
uv run python mcp_stdio.py
MCP Tool: youtube_search
Tool Name: youtube_search
Description: Search YouTube videos and return complete metadata
Parameters:
| Parameter | Type | Required | Default | Description |
|---|---|---|---|---|
keyword |
string | ✅ | - | Search keyword (1-200 characters) |
limit |
integer | ❌ | 1 | Number of results to return (1-100) |
sort_by |
string | ❌ | relevance | Sort order: relevance or date |
Response Format:
{
"videos": [
{
"video_id": "mIF-nn_y2_8",
"title": "Jackie Cheung - Farewell Kiss",
"channel": "Music Without Boundaries",
"url": "https://www.youtube.com/watch?v=mIF-nn_y2_8",
"channel_url": "https://www.youtube.com/@channel_name",
"publish_date": "2020-01-15",
"view_count": "1500000",
"description": "Classic Cantonese song..."
}
],
"message": "Successfully returned 1 result"
}
Error Handling:
- Empty keyword: Returns
{"error": "INVALID_KEYWORD", "message": "Search keyword cannot be empty..."} - Invalid limit: Returns
{"error": "INVALID_LIMIT", "message": "limit must be between 1-100..."} - YouTube service unavailable: Returns
{"error": "SERVICE_UNAVAILABLE", "message": "YouTube service is temporarily unavailable..."} - Cache service failure: Gracefully degrades to direct search and returns normal results
Environment Configuration
Copy .env.example to .env and modify as needed:
cp .env.example .env
Docker Deployment
docker-compose up -d
Testing
pytest tests/
Project Structure
src/youtube_search/
├── models/ # Pydantic data models
├── services/ # Business logic layer
├── api/ # API routes
└── utils/ # Utility functions
License
MIT License
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.