youtube

This MCP server exposes the YouTube Data API v3, Analytics API, and Reporting API through the Model Context Protocol, giving your AI assistants full access to your YouTube channel for analytics, video management, comments, and more. It bundles a suite of around 40 tools, including channel and video data retrieval, analytics metrics (watch time, retention, audience insights), bulk reporting, video upload and management, playlist actions, caption handling, transcript extraction, and SEO-related features like autocomplete suggestions and trending data. The server coordinates three YouTube APIs and provides a unified command surface so you can query channel data, publish content, post comments, and generate reports in a structured, tool-oriented way.

To use it, install the MCP server (Python-based) and configure your environment with OAuth credentials and API keys as needed. On first run, you will go through an OAuth consent flow in your browser. After authentication, the server stores tokens securely on disk and refreshes them automatically. You can then invoke the available tools (for example youtube_get_channel, youtube_upload_video, youtube_analytics_overview, youtube_reporting_list_jobs, youtube_list_comments, youtube_search, youtube_trending, and many more) via your MCP client configuration to perform tasks or answer questions that require access to your YouTube data.

Prerequisites:

• Python 3.11+ (recommended)
• pip (comes with Python)
• A Google Cloud project with YouTube Data API v3, YouTube Analytics API, and YouTube Reporting API enabled
• OAuth 2.0 client credentials (downloaded as client_secret.json) or set YOUTUBE_MCP_CLIENT_SECRET to the path of your credentials

Install the MCP server package:

# Install the MCP server package
pip install youtube-studio-mcp

Optionally, install with uv (if you prefer an isolated environment manager):

uv pip install youtube-studio-mcp

Configure credentials (example using environment variable):

export YOUTUBE_MCP_CLIENT_SECRET=/path/to/client_secret.json
# Optional: your public API key for non-auth data
export YOUTUBE_API_KEY=your-api-key-for-public-data

Run the MCP server:

python -m youtube_studio_mcp

If you prefer to run via uvx (for Python environments):

uvx youtube-studio-mcp

First run will prompt you to authorize access in a browser; tokens are saved to a default location (e.g., ~/.youtube-mcp/token.json) and refreshed automatically.

Tips and notes:

• The server tracks quota usage on the client side and will hard-fail when you exhaust your quota.
• Use youtube_auth_status to check current quota usage and authentication status.
• If you rotate credentials, update YOUTUBE_MCP_CLIENT_SECRET so the MCP server can re-authorize.
• The environment supports both YouTube Data API read operations and write operations (video upload, metadata updates, comments, etc.). Be mindful of quota costs for heavy tasks like bulk reporting or video uploads.
• For production use, consider securing the OAuth client secret file and restricting access to the token storage location (e.g., ~/.youtube-mcp).
• If you need to test specific capabilities, consult the Tools Reference in the README to discover exact tool names and parameters for channel, analytics, publishing, and bulk reporting tasks.