wyoming-protocol

Speech-To-Text and Text-To-Speech component for Home Assistant using Apple's on-device STT and TTS models and Wyoming protocol. Great performance, full privacy, always available

A Wyoming protocol ASR proxy that verifies speaker identity and isolates voice commands from background noise before forwarding audio to a downstream speech-to-text service. Designed for Home Assistant voice pipelines to prevent false activations from TVs, radios, and other people - and to deliver clean transcripts even in noisy environments.

Open-source, fully offline wake word / keyword spotting (KWS) inference engine — ONNX/TFLite <130KB, multi-keyword, multilingual. Train a custom wake word online with auto hard-negative mining; deploy on ESP32 / Raspberry Pi / Android / Web / Windows / macOS / Linux / Home Assistant — runs on any ONNX/TFLite-capable device.

Wyoming Protocol TTS and STT & MLX LLM server for iOS/macOS

GPU-accelerated Qwen3-TTS WebUI and OpenAI-compatible API with integrated Wyoming TTS for Docker and Unraid

Drop-in Amiga narrator.device replacement with modern neural voices — forwards text to a Piper TTS server over the Wyoming protocol, plays via AHI.

🗣 Verify speaker identity and clean voice audio for accurate speech-to-text in Home Assistant using Wyoming protocol ASR proxy.

silero-api-server

Mobile voice satellite for Home Assistant and AI agents, running on the Pollen Robotics Microduck

Kratt: Estonian Home Assistant voice prototype

Wyoming protocol gateway for Home Assistant with AI-powered voice processing

Docker image to make index tts run with Home Assistant

Fully local Home Assistant voice pipeline (Wyoming Whisper + Piper) — Proxmox/x86 and Apple Silicon deployment templates

Wyoming Protocol to Realtime Speech-to-Speech API bridge for Home Assistant

Async Python orchestrator that turns a Wyoming voice satellite into a fully-local voice assistant driven by any OpenAI-compatible LLM.

Home Assistant OS add-ons: Parakeet TDT 0.6B v3 speech-to-text on the Intel NPU, and Silero v5 Russian text-to-speech on the CPU — both over the Wyoming protocol, built from source, every download SHA-256 pinned.

Hermes Agent plugin for text-to-speach using a remote piper server over the wyoming protocol

WIP: Wyoming protocol wrapper for audio.cpp

GigaAM-v3 speech recognition integration for Home Assistant via Wyoming protocol.

Retrotalk: a retro formant text-to-speech engine, written from scratch, as a Home Assistant Wyoming add-on (work in progress)