VideoGraphAI 🎬

• 🌟 Overview
• ✨ Key Features
• 🔄 Workflow
• 📋 Prerequisites
• 🚀 Installation
• ⚙️ Configuration
• 📝 Usage
• 🔧 Troubleshooting
• 👥 Contributing
• 🙏 Acknowledgements
• 📄 License

VideoGraphAI streamlines the creation of YouTube Shorts using advanced AI technologies. Built with Streamlit, it offers end-to-end video production capabilities from content research to final rendering. The system leverages various AI models and APIs to create engaging, relevant content automatically.

• 🔍 Real-time Research: Automated content research using Tavily Search API
• 📝 AI Script Generation: Flexible LLM compatibility (OpenAI, Groq, etc.)
• 🎨 Dynamic Visuals: Image generation via TogetherAI (FLUX.schnell)
• 🎤 Professional Audio: Voiceovers using F5-TTS
• 📺 Automated Subtitles: Synchronized captions with Gentle
• 🖥️ User-Friendly Interface: Built with Streamlit for easy operation

1. Input → User provides topic, timeframe, and video length
2. Research → AI researches recent events using graph agents
3. Content Creation → Generates titles, descriptions, hashtags, and script
4. Media Production → Creates storyboard and acquires media assets
5. Audio & Subtitles → Generates voiceover and synchronized captions
6. Compilation → Assembles final video with all components
7. Delivery → Presents downloadable video through Streamlit interface

• Python 3.8+
• FFmpeg
• Docker (optional, recommended for Gentle server)
• API Keys:
  • Groq API
  • Together AI API
  • Tavily Search API
  • F5-TTS (local installation)

git clone https://github.com/mikeoller82/VideoGraphAI.git
cd VideoGraphAI

# Option 1: Conda (Recommended)
conda create -n videographai python=3.8 pip
conda activate videographai

# Option 2: Virtual Environment
python3 -m venv venv
source venv/bin/activate  # Windows: venv\Scripts\activate

pip install -r requirements.txt

sudo apt update
sudo apt install ffmpeg

brew install ffmpeg

git clone https://github.com/SWivid/F5-TTS.git
cd F5-TTS
pip install -r requirements.txt
# Follow F5-TTS documentation for torch and CUDA setup then all you do is take your sample wav file and put it in /F5-TTS/src/f5_tts/infer/examples/basic
#  this will be inside your VideoGraphAI directory after you git clone it. then just config toml file either inside F5 there is a basic.toml file but in the functiion
#  genrate_voicoever is the override toml that will be used so you can just configure it there . Honestly its whatever you prefer then youl be good to go on any voicoever you want just need the a wav file of like 5 seconds to 8 seconds minimal
cd ..

docker run -d -p 8765:8765 lowerquality/gentle

Create a .env file:

GROQ_API_KEY=your_groq_api_key
BFL_API_KEY=your_black_forest_labs_api_key
TOGETHER_API_KEY=your_together_api_key
TAVILY_API_KEY=your_tavily_api_key

1. Launch application:

streamlit run app.py

2. Enter parameters:

   • Topic for your video
   • Time frame (past month/year/all)
   • Video length (60/120/180 seconds)

3. Click "Generate Video" and wait for processing

We welcome contributions! Please follow these steps:

1. Fork the repository
2. Create your feature branch (git checkout -b feature/AmazingFeature)
3. Commit changes (git commit -m 'Add AmazingFeature')
4. Push to branch (git push origin feature/AmazingFeature)
5. Open a Pull Request

• F5-TTS: Advanced text-to-speech capabilities

@article{chen-etal-2024-f5tts,
    title={F5-TTS: A Fairytaler that Fakes Fluent and Faithful Speech with Flow Matching},
    author={Chen, Yushen and Niu, Zhikang and Ma, Ziyang and Deng, Keqi and Wang, Chunhui and Zhao, Jian and Yu, Kai and Chen, Xie},
    journal={arXiv preprint arXiv:2410.06885},
    year={2024}
}

• TogetherAI: Image generation via FLUX.schnell model (https://www.together.ai/)

• Tavily: Real-time search capabilities (https://tavily.com/#api)

This project is licensed under the MIT License - see the LICENSE file for details.