Omnix Studio

Omnix is a local multi-modal AI studio that allows you to orchestrate vision, speech, and text models entirely on your machine. It also provides a robust local API for other applications to use Omnix as an inference engine.

Features

• Multi-Modal: Support for Text, Vision, STT, TTS, Image Generation, and Music Generation.
• Local First: All models run locally using WebGPU or WASM.
• Theme Support: Polished Light and Dark modes.
• Live Mode: Real-time screen and voice analysis.
• Sandbox: Built-in environment for generating and running code.

---

Local API Guide

Omnix provides a local API running on http://localhost:3000/api.

Endpoints

1. Text Generation (POST /api/text)

• Body: {"prompt": "string", "systemPrompt": "string"}
• Response: {"response": "string"}

2. Vision Analysis (POST /api/vision)

• Body: multipart/form-data
  • image: File (Binary)
  • prompt: string (Optional)
• Response: {"caption": "string", "response": "string"}

3. Director Routing (POST /api/director)

• Body: {"prompt": "string"}
• Response: {"intent": "string", "prompt": "string"}

4. Image Generation (POST /api/image)

• Body: {"prompt": "string"}
• Response: {"status": "success", "url": "string"}

5. Music Generation (POST /api/music)

• Body: {"prompt": "string"}
• Response: {"status": "success", "audioUrl": "string"}

6. Speech-to-Text (POST /api/stt)

• Body: multipart/form-data
  • audio: File (WAV/MP3)
• Response: {"text": "string"}

7. Text-to-Speech (POST /api/tts)

• Body: {"text": "string", "voice": "string"}
• Response: {"status": "success", "audioUrl": "string"}

Example Usage (CURL)

curl -X POST http://localhost:3000/api/text \
     -H "Content-Type: application/json" \
     -d '{"prompt": "Hello Omnix!"}'

---

Electron Setup Guide

The desktop version of Omnix provides unrestricted RAM access, WebGPU acceleration, and native filesystem integration.

Precompiled

Prerequisites

• Node.js: v18 or higher recommended.
• NPM: Standard package manager.

Installation

1. Clone the repository (if you haven't already):

   git clone https://github.com/LoanLemon/Omnix
   cd omnix

2. Install dependencies:

   npm install

Running the Application

Development Mode

To run the app in development mode with hot-reloading:

1. Start the Vite development server:

   npm run start

Production Build

To package the application for production:

1. Build the app:

   # Note: You may need to install electron-builder or electron-packager for full distribution
   npm run electron:build

Desktop Features

• Unrestricted RAM: Up to 16GB of heap memory for large models.
• WebGPU Acceleration: Hardware acceleration enabled by default.
• Minimize to Tray: Moves to system tray on close/minimize.
• Local Filesystem: Direct interaction with local files.

Troubleshooting

• WebGPU Errors: Ensure your graphics drivers are up to date. Some older GPUs may not support WebGPU.
• Port Conflicts: If port 3000 is occupied, the Electron app may fail to connect in dev mode.

---

Developed by Dustin Lee at LemOne Labs.