llm-youtube-transcript 0.1

Plugin for LLM that adds support for fetching transcripts from YouTube videos using Supadata API

llm-youtube-transcript

LLM plugin for fetching YouTube video transcripts.

This plugin fetches the text transcript of a YouTube video using the Supadata API and makes it available as an LLM fragment.

For background on LLM fragments:

• Fragments - LLM Documentation

Installation

Install this plugin in the same environment as LLM.

llm install llm-youtube-transcript

Configuration

This plugin requires a Supadata API key to function. You get 100 requests per month with a free acount.

1. Sign up at Supadata.ai and get your API key.

2. Set the API key as an environment variable named SUPADATA_API_KEY:

   export SUPADATA_API_KEY=your_api_key_here
   

   You can add this line to your shell profile (e.g., ~/.bashrc, ~/.zshrc) to make it persistent across sessions.

Usage

You can feed the transcript of a YouTube video into LLM using the yt: fragment prefix followed by the video URL.

Make sure you have set the SUPADATA_API_KEY environment variable first.

Note: If your YouTube URL contains special characters (like ? or &), you must enclose the entire fragment identifier (e.g., yt:https://...) in single or double quotes on the command line. This prevents your shell from misinterpreting the URL.

For example:

# Fetch transcript from a YouTube video URL (note the quotes)
llm -f 'yt:https://www.youtube.com/watch?v=dQw4w9WgXcQ' 'Summarize this transcript'

# Example with a different video
llm -f 'yt:https://www.youtube.com/watch?v=VIDEO_ID_HERE' 'What are the main points discussed in this video?'

The plugin will call the Supadata API to retrieve the plain text transcript of the specified video. If the API key is missing or invalid, or if the transcript cannot be fetched, an error will be raised.

Development

To set up this plugin locally, first checkout the code. Then create a new virtual environment:

cd llm-youtube-transcript
python -m venv venv
source venv/bin/activate

Now install the dependencies and test dependencies:

# Installs llm, httpx, and testing tools like pytest
pip install -e '.[test]'

To run the tests:

python -m pytest