FastAPI voice-assistant framework for prototyping LLM-powered voice assistants in minutes
Nala is a FastAPI-based voice-assistant framework that integrates speech-to-text, text-to-speech, and LLMs to prototype voice assistants in under five minutes. It provides a browser front-end using the Web Speech API, an audio-to-audio FastAPI backend, and helper functions to plug in models like Whisper, SpeechT5/Bark, and Dolly. It is suited for developers building web or device voice assistants and experimenting with conversational audio workflows.
0