Speech To Speech: Build voice agents with open-source models
Summary
The repository 'Speech To Speech' describes a low-latency, modular voice-agent pipeline (VAD -> STT -> LLM -> TTS) with an OpenAI Realtime-compatible WebSocket API. It supports multiple backends for STT, LLM, and TTS, including local, self-hosted, and provider options, and provides detailed installation, run modes, and multilingual support information. The project serves as a practical open-source platform for building speech-enabled AI applications and demos its adoption in real-world robots like Reachy Mini.