🎬 ClippedAI: Autonomous YouTube Shorts Generator

🔴 Live Demo: https://clippedai.onrender.com

An end-to-end, production-ready SaaS application that autonomously transforms long-form videos into viral YouTube Shorts using advanced acoustic heuristics, OpenAI Whisper, and Groq's LLaMA3.

✨ Features

🧠 Algorithmic Hook Detection: Uses acoustic analysis (energy, RMS, WPM) combined with Groq LLaMA3 to automatically isolate high-dopamine, viral moments from hours of raw footage.
⚡ Cloud AI Pipeline: Compresses and streams audio to OpenAI Whisper API for blazing-fast, word-level timestamps.
🎨 Kinetic Typography: Automatically renders highly-engaging, animated captions using ffmpeg and moviepy that scale and pop to retain viewer attention.
🔐 Google OAuth 2.0: Secure user authentication and dashboard gating.
💳 Razorpay Monetization: Fully integrated payment gateway to purchase generation credits.
📺 YouTube Auto-Publishing: Users can link their YouTube channels to automatically render and broadcast shorts while they sleep.
🛠️ Dev-Mode Fallbacks: Seamlessly bypasses Google Auth and Razorpay when testing locally without API keys!

🏗️ Architecture

graph TD
    A[User Uploads Video] --> B(Local Acoustic Analysis)
    B --> C{Extract Audio}
    C --> D[OpenAI Whisper API]
    D --> E[Word-Level Timestamps]
    B --> F[Groq LLaMA3]
    F --> G[Viral Hook Ranking & Titles]
    E --> H(Kinetic Subtitle Rendering)
    G --> H
    H --> I[Final 4K Vertical Short]
    I --> J[YouTube Auto-Publish]

🚀 Quick Start (Local Development)

We've built a Dev-Mode Bypass so you can test the entire dashboard and generation pipeline immediately, even before you set up your API keys!

1. Clone & Install

git clone https://github.com/YourUsername/ClippedAI.git
cd ClippedAI

# Create virtual environment
python -m venv venv
source venv/bin/activate  # Or `.\venv\Scripts\activate` on Windows

# Install dependencies
pip install -r requirements.txt

2. Environment Variables

Copy the template to create your .env file:

cp .env.template .env

(Leave the dummy keys as they are to test the Dev-Mode bypass, or insert your real keys to use production services).

3. Run the App

python app.py

Visit http://localhost:5000. Click Login, and the system will instantly drop you into the Founder dashboard!

☁️ Production Deployment (Render)

This repository includes a production-ready Dockerfile optimized to run the heavy AI pipeline on extremely low memory (fitting within Render's Free Tier!).

Go to Render.com and create a New Web Service.
Connect this GitHub repository.
Add your real API keys as Environment Variables.
Deploy! Render will automatically install Python, ffmpeg, and Gunicorn using the Dockerfile.

📦 Tech Stack

Backend: Python, Flask, SQLAlchemy, Gunicorn
Frontend: Vanilla JS, CSS Glassmorphism UI
Video Processing: FFmpeg, MoviePy
AI Models: OpenAI Whisper (Transcription), Groq LLaMA3 (Hook Detection)
Integrations: Google Cloud OAuth, Razorpay, YouTube Data API v3

Recording.2026-06-08.171646.1.mp4

📝 License

Distributed under the MIT License. See LICENSE for more information.

FilesExpand file tree

README.md

Latest commit

History