# Klavis-AI/YouTube [Health: Active]

**Category:** 🌐 Social Media  
**Repository:** https://github.com/Klavis-AI/klavis/tree/main/mcp_servers/youtube  
**GitHub Stars:** 5827  
**Views:** 4  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/klavis-ai-youtube

## Description
Extract and convert YouTube video information.

## Claude Desktop Quick Installation
Install path detected from listing signals. Uses `uvx` (confidence: high):

```json
"mcpServers": {
  "youtube": {
    "command": "uvx",
    "args": ["klavis"],
    "env": {
      "API_KEY": ""
    }
  }
}
```

**Requires environment variables:** `API_KEY` — the values above are empty placeholders; fill in real credentials before running (see the repository for what each one is for).

## Documentation

## What Klavis-AI/YouTube does

Klavis-AI/YouTube connects MCP clients and agent applications with YouTube data. Its tool set covers several common retrieval tasks: obtaining video transcripts with timestamps, reading video details and metadata, searching for videos, inspecting channels and their video listings, and accessing playlist contents and metadata.

The server is suited to workflows that need YouTube information in an agent-readable form. For example, an agent can locate videos by keywords, gather a video's title and description, retrieve available transcript text, or inspect the videos associated with a channel or playlist.

## How it works

Requests are handled through the YouTube Data API. The server can be accessed through Klavis's hosted infrastructure or run as a container for self-hosting. The hosted route is initialized through the Klavis SDK: applications create a `Klavis` client with a Klavis API key and then request a YouTube server instance for a user.

For local or controlled deployments, the published Docker image runs the YouTube MCP server and maps its service to port 5000. The container receives an API key through an environment variable and uses a YouTube Data API key configured through Google Cloud.

## Setup and configuration

The hosted option requires installing the Klavis package for Python or npm and obtaining a Klavis API key. The server instance is created by selecting the `YOUTUBE` service and providing a user identifier.

Self-hosting requires Docker and the image `ghcr.io/klavis-ai/youtube-mcp-server:latest`. Start the container with port 5000 exposed and pass the API key through `API_KEY`:

```bash
docker pull ghcr.io/klavis-ai/youtube-mcp-server:latest
docker run -p 5000:5000 -e API_KEY=$API_KEY \
  ghcr.io/klavis-ai/youtube-mcp-server:latest
```

Create the YouTube credential in Google Cloud Console and enable YouTube Data API v3 before using the self-hosted deployment.

## Tools and capabilities

- Retrieve full video transcripts, including timestamps.
- Read video details such as titles, descriptions, and statistics.
- Search YouTube videos using keywords and filters.
- Get channel details and lists of channel videos.
- Access playlist contents and playlist metadata.

## Limitations and notes

The server depends on YouTube Data API v3, so the relevant Google Cloud API must be enabled for self-hosted use. Hosted deployments use Klavis infrastructure and require a Klavis API key. The repository is licensed under Apache 2.0.

_Full upstream README: https://allmcps.com/mcp/klavis-ai-youtube/readme_