iWish: AI-Powered YouTube Transcription & Summarization

iWish is a streamlined CLI tool that transforms YouTube videos into organized text data. It uses OpenAI's Whisper for high-accuracy transcription and OpenRouter (Gemini) for intelligent summarization.

📦 Installation

1. Requirements

• Python 3.8+
• ffmpeg installed in your system.

2. Install

The easiest way is to install directly via pip:

pip install git+https://github.com/BroKarim/iwish.git

🛠 Usage

Simply provide the YouTube URL:

iwish "https://www.youtube.com/watch?v=dQw4w9WgXcQ"

Output Structure

iWish automatically creates a folder named after the video title:

Title_of_the_Video/
├── subtitle.txt    <-- The full transcription
└── summary.md     <-- AI-generated structured summary

Options

# Use a more accurate model (tiny, base, small, medium, large, turbo)
iwish "URL" --model medium

# Output in different format (vtt, srt, txt)
iwish "URL" --format vtt

# Disable summarization (enabled by default)
iwish "URL" --no-summarize

# Translate foreign audio to English
iwish "URL" --task translate

🤖 AI Summarization

On your first run with summarization enabled, iWish will ask for your OpenRouter API key.

• It is saved locally in ~/.iwish_config.json.
• You don't need to enter it again.
• Default model: gemini-2.0-flash.

📄 License

This project is open-source and licensed under the MIT License.