home / mcp / yt-dlp mcp server
A Model Context Protocol (MCP) server that bridges Video & Audio content with Large Language Models using yt-dlp.
Configuration
View docs{
"mcpServers": {
"kevinwatt-yt-dlp-mcp": {
"command": "npx",
"args": [
"-y",
"@kevinwatt/yt-dlp-mcp@latest"
],
"env": {
"YTDLP_COOKIES_FILE": "/path/to/cookies.txt",
"YTDLP_DOWNLOADS_DIR": "/path/to/downloads",
"YTDLP_CHARACTER_LIMIT": "25000",
"YTDLP_DEFAULT_RESOLUTION": "1080p",
"YTDLP_COOKIES_FROM_BROWSER": "chrome",
"YTDLP_DEFAULT_SUBTITLE_LANG": "en",
"YTDLP_MAX_TRANSCRIPT_LENGTH": "50000"
}
}
}
}You can run yt-dlp with MCP-compatible agents to search, fetch metadata, download videos or audio, and extract transcripts and subtitles. This server is designed to be easy to drop into MCP workflows so your AI agents can interact with video content across platforms with safety and structure.
You connect an MCP client to the yt-dlp MCP server to perform capabilities like searching, fetching video metadata, downloading videos or audio, and obtaining transcripts or subtitles. Use natural language prompts to trigger tools such as searching for videos, listing available subtitles, downloading content in a chosen format, or getting a clean metadata summary. Tools are designed to be intuitive and predictable, returning structured data or files that your agent can process.
Prerequisites you need before starting:
Install yt-dlp on your operating system using the following commands.
# Windows
winget install yt-dlp
```
```bash
# macOS
brew install yt-dlp
```
```bash
# Linux
pip install yt-dlpInstall the MCP server as a stdio process and connect it to your MCP client using the provided configuration.
{
"mcpServers": {
"yt-dlp": {
"command": "npx",
"args": ["-y", "@kevinwatt/yt-dlp-mcp"]
}
}
}If you prefer installing globally and running directly, use the following command.
npm install -g @kevinwatt/yt-dlp-mcpSearch YouTube with pagination and date filtering, output JSON or Markdown, and filter by relevance and quality.
List all available subtitle languages for a video, including formats and auto-generated status.
Download subtitles in VTT format with timestamps for a given video and optional language.
Generate a clean plain text transcript for a video with optional language.
Download video to the Downloads folder with optional start and end times and resolution control.
Extract and download the best quality audio (M4A/MP3) from a video.
Extract comprehensive video metadata in JSON with optional field filtering.
Get a human-readable metadata summary for quick reference.