---

VT.ai

Minimal multimodal AI chat app with dynamic conversation routing

🚀 Features

Multi-Provider AI Orchestration

✅ Supported Models Provider:

• DeepSeek
• OpenAI
• Anthropic
• Google Gemini
• Local Models via Ollama (Llama3, Phi-3, Mistral, etc.)
• Cohere
• OpenRouter

✨ Core Capabilities:

• Dynamic conversation routing with SemanticRouter
• Multi-modal interactions (Text/Image/Audio)
• Assistant framework with code interpretation
• Real-time response streaming
• Cross-provider model switching
• Local model support with Ollama integration

📦 Installation

Prerequisites

• Python 3.11+ (specified in .python-version)
• uv (for managing virtual environments and dependencies)
• Ollama (for local models, optional)

Install uv if not already installed

pip install uv
git clone https://github.com/vinhnx/VT.ai.git
cd VT.ai

Create a virtual environment and install dependencies with uv

uv venv
source .venv/bin/activate  # Linux/Mac
# .venv\Scripts\activate     # Windows
uv pip install -r requirements.txt
cp .env.example .env

Note: uv provides a fast and efficient way to manage Python environments and dependencies. Activate the virtual environment before running the app.

🔧 Configuration

Populate the .env file with your API keys. Depending on the models and features you want to use, set the corresponding keys. For example:

• OPENAI_API_KEY: Required for OpenAI models, assistant mode, TTS, and image generation.
• GEMINI_API_KEY: Required for Google Gemini models and vision capabilities.
• ANTHROPIC_API_KEY: Required for Anthropic models.
• GROQ_API_KEY: Required for Groq models.
• COHERE_API_KEY: Required for Cohere models.
• OPENROUTER_API_KEY: Required for OpenRouter models.
• MISTRAL_API_KEY: Required for Mistral models.
• HUGGINGFACE_API_KEY: Required for Hugging Face models (if applicable).

Refer to .env.example for the full list. Example:

OPENAI_API_KEY=sk-your-key
GEMINI_API_KEY=your-gemini-key
COHERE_API_KEY=your-cohere-key
ANTHROPIC_API_KEY=your-claude-key
HUGGINGFACE_API_KEY=your-huggingface-key
GROQ_API_KEY=your-groq-key
OPENROUTER_API_KEY=your-openrouter-key
MISTRAL_API_KEY=your-mistral-key

# For local models via Ollama
OLLAMA_HOST=http://localhost:11434

Note: If OPENAI_API_KEY or GEMINI_API_KEY is missing, the app will prompt you to enter them at startup.

🖥️ Usage

Start Application

Activate the virtual environment

source .venv/bin/activate  # Linux/Mac
# .venv\Scripts\activate     # Windows

(Optional) Train semantic router for customization

Requires OPENAI_API_KEY

python src/router/trainer.py

Launch interface

chainlit run src/app.py -w

Note: Training the semantic router is optional. Use the provided layers.json for default routing or train your own for better performance/customization. Training requires an OpenAI API key.

Key Commands

Shortcut	Action
Ctrl+/	Switch model provider
Ctrl+,	Open settings
Ctrl+L	Clear conversation history

🧩 Chat Profiles

Standard Chat Mode

• Multi-LLM conversations
• Dynamic model switching
• Image generation & analysis
• Audio transcription

Assistant Mode (Beta)

Example assistant capabilities

async def solve_math_problem(problem: str):
    assistant = MinoAssistant()
    return await assistant.solve(problem)

• Code interpreter for complex calculations
• File attachments (PDF/CSV/Images)
• Persistent conversation threads
• Custom tool integrations

🏗️ Project Structure

VT.ai/
├── src/
│   ├── assistants/       # Custom AI assistant implementations
│   ├── router/           # Semantic routing configuration
│   ├── utils/            # Helper functions & configs
│   └── app.py            # Main application entrypoint
├── public/               # Static assets
├── requirements.txt      # Python dependencies
└── .env.example          # Environment template

🌐 Supported Models

Category	Models
Chat	GPT-4o, Claude 3.5, Gemini 1.5, Llama3-70B, Mixtral 8x7B
Vision	GPT-4o, Gemini 1.5 Pro, Llama3.2 Vision
Image Gen	DALL-E 3
TTS	OpenAI TTS-1, TTS-1-HD
Local	Llama3, Phi-3, Mistral, Deepseek R1 series

🤝 Contributing

Development Setup

# Activate the virtual environment
source .venv/bin/activate  # Linux/Mac
# .venv\Scripts\activate     # Windows

Install development tools (e.g., for testing or formatting)

uv pip install pytest black

# Run tests (once tests are added)
pytest tests/

# Format code
black .

Contribution Guidelines

1. Fork the repository
2. Create a feature branch (git checkout -b feature/amazing-feature)
3. Add type hints for new functions
4. Update documentation
5. Open a Pull Request

Note: Development dependencies (e.g., pytest, black) are not currently specified in a separate file. Install them manually with uv pip install as needed.

📄 License

MIT License - See LICENSE for full text.

🌟 Acknowledgements

• Inspired by Chainlit for chat interface
• Powered by LiteLLM for model abstraction
• Semantic routing via SemanticRouter