Real-Time AI Voice Chat

Have natural conversations with AI! This project allows you to chat with large language models (LLMs) using just your voice, with near real-time spoken responses. Think of it as your own digital conversation partner. What’s under the hood? A sophisticated client-server system for low-latency interaction: - **Capture**: Your voice is recorded by the browser. - **Stream**: Audio chunks are sent to a Python backend via WebSockets. - **Transcribe**: RealtimeSTT quickly converts your speech to text. - **Think**: The text is sent to an LLM (e.g., Ollama or OpenAI) for processing. - **Synthesize**: Using RealtimeTTS, the AI’s text response is turned back into speech. - **Return**: The generated audio streams back to your browser for playback. - **Interrupt**: Jump in anytime! The system gracefully handles interruptions. Key Features: - **Fluid Dialogue**: Speak and listen as in a real chat. - **Live Feedback**: See partial transcriptions and AI responses in real time. - **Low-Latency Focus**: Optimized architecture using audio chunk streaming. - **Smart Turn-Taking**: Dynamic silence detection (turndetect.py) adapts to conversation flow. - **Flexible AI Brain**: Plug-and-play LLM backends (Ollama by default, OpenAI supported via llm_module.py). - **Customizable Voices**: Choose from different text-to-speech engines (via audio_module.py). - **Web Interface**: Clean and simple UI built with vanilla JS and the Web Audio API. - **Dockerized Deployment**: Recommended setup with Docker Compose for easy dependency management.

Real-Time AI Voice Chat lets users converse naturally with LLMs via voice, receiving real-time responses as a digital companion.