An open API service indexing awesome lists of open source software.

https://github.com/PrajwalPrashanth/youtube-mcp-server

MCP server to fetch and extract transcripts from YouTube videos. Enable AI/LLMs to have access to transcripts for further actions like summarize, write blog..
https://github.com/PrajwalPrashanth/youtube-mcp-server

claude mcp mcp-server model-context-protocol transcripts youtube

Last synced: 2 months ago
JSON representation

MCP server to fetch and extract transcripts from YouTube videos. Enable AI/LLMs to have access to transcripts for further actions like summarize, write blog..

Awesome Lists containing this project

README

          

# YouTube MCP(Model Context Protcol) Server

----
This MCP server provides a tool to fetch and extract transcripts from YouTube videos. It enables AI language models (LLMs) to:

- Access complete text transcripts from YouTube video URLs
- Process and analyze video content through transcripts
- Reference and discuss video information in conversations

The server acts as a bridge between AI models and YouTube content by automatically extracting and formatting video transcripts.

Instead of saying can't access links
![](image.png)

## Setup
1. Install uv (Python package manager):

**Mac:**
```bash
brew install uv
```

**Linux:**
```bash
curl -LsSf https://astral.sh/uv/install.sh | sh
```

**Windows:**
```powershell
powershell -c "irm https://astral.sh/uv/install.ps1 | iex"
```

2. Clone this repository:
```bash
git clone git@github.com:PrajwalPrashanth/youtube-mcp-server.git
cd youtube-mcp-server
```

3. Create a virtual environment and install dependencies:
```bash
uv venv
source .venv/bin/activate
# On Windows: .venv\Scripts\activate
uv pip install -r pyproject.toml
```

## Add this MCP server to Claude Desktop ([download](https://claude.ai/download))
```bash
uv run mcp install -e . server.py -n "youtube-mcp"
```

Once this is done you should see the tool icon in the UI
![](mcp-avail.png)

and you can ask for it to do actions (like summarize) with the transcript by providing a YT video

![](allow-tool.png)