clipperr 🚀

AI-Powered Viral Video Clipper.
Cut long-form videos into viral short-form clips (TikTok/Reels/Shorts) with automatic face tracking, speaker diarization, and AI-driven viral analysis—all running locally on your hardware.

---

🌌 Why clipperr?

• 🔒 Privacy First: Your videos stay on your machine. No cloud uploads required.
• 💰 Zero Cost: High-end AI processing without monthly subscription fees.
• 🤖 Intelligent Focus: Automatically keeps the active speaker centered using advanced face tracking and speaker diarization.
• ⚡ High Performance: Ultra-fast rendering engine powered by Rust and FFmpeg.

---

🛠️ Architecture

clipperr combines a modern Python/PySide6 frontend with a high-performance Rust core engine.

graph TB
    subgraph UI ["Desktop UI (PySide6)"]
        Main["Dashboard"]
        Review["Review & Preview"]
        Settings["AI Model Manager"]
    end

    subgraph Pipeline ["AI Processing Pipeline"]
        Whisper["Whisper (Transcription)"]
        Diarization["Pyannote (Speaker ID)"]
        LLM["Gemma 3 (Viral Analysis)"]
        Tracking["Face Tracking Engine"]
    end

    subgraph Core ["System Core"]
        Rust["Rust PyO3 Bridge"]
        FFmpeg["FFmpeg Video Pipeline"]
    end

    Main --> Pipeline
    Pipeline --> Rust
    Rust --> FFmpeg
    FFmpeg --> Main

Loading

---

🚀 Getting Started

📦 Windows (Recommended)

1. Download the latest Windows Release (clipperr-windows-v1.0.zip).
2. Extract the folder and run clipperr.exe.
3. Go to Settings ⚙️ and paste your HuggingFace Access Token.
4. Start clipping!

🐧 Linux (From Source)

Ensure you have ffmpeg and python 3.12+ installed.

git clone https://github.com/rasyiqi-code/clipperr.git
cd clipperr
pip install -r requirements.txt
python3 app/main.py

---

🧬 Technical Flow

sequenceDiagram
    participant U as User
    participant P as Pipeline
    participant AI as AI Models
    participant R as Rust Core

    U->>P: Upload Video
    P->>AI: Transcribe & Identify Speakers
    P->>AI: Analyze for Viral Moments
    loop Each Clip
        P->>AI: Auto-Track Active Speaker
        P->>R: Optimized FFmpeg Rendering
    end
    P-->>U: Ready to Share!

Loading

---

⚙️ Requirements

• OS: Windows 10+ or Modern Linux.
• RAM: 8GB Minimum (16GB+ recommended for faster processing and stability).
• GPU: NVIDIA GPU (Optional, for 5x faster processing).
• Disk: ~5GB free space for AI models.

---

📄 License

MIT License - Developed for creators who value privacy and power.