ViGen

ViGen is a desktop application powered by AI that enables users to generate videos from simple ideas or detailed scripts. It leverages Google's Gemini models for story and script generation, and Veo for video synthesis, wrapped in a modern PyQt6 interface.

Features

• Idea to Video: Generate a full story, character portraits, script, and final video from a single idea.
• Script to Video: Input a script and character details to produce a video directly.
• Multimodal AI: Uses Google Gemini for text/image logical processing and Veo for video generation.
• Character Consistency: Maintains character appearance across scenes using reference images.
• Cross-Platform: Available for Windows and macOS (Intel & Apple Silicon).

Installation

Download Release

Go to the Releases page and download the executable for your operating system:

• Windows: ViGen_Windows.exe
• macOS: ViGen_macOS_Silicon.zip (or Intel)

Run from Source

1. Clone the repository.
2. Install dependencies:

   pip install -r requirements.txt

   (Or use uv / poetry as configured in the project)
3. Run the application:

   python -m gui.main

Configuration

Upon first launch, ViGen will create a configs folder with template files. You must provide your own API keys for the application to function.

1. Open configs/idea2video.yaml and configs/script2video.yaml.
2. Enter your keys:
   • chat_model: OpenRouter API Key (or other supported provider).
   • image_generator: Google Cloud/Gemini API Key.
   • video_generator: Google Cloud/Gemini API Key.

Technology Stack

• GUI: PyQt6, Qt-Material
• AI Orchestration: LangChain
• Models: Google Gemini 2.5 Flash Image, Imagen 3, Veo 3.1
• Video Processing: MoviePy

License

This project is licensed under the MIT License.