Faster-Qwen3-TTS-Wyoming ghcr.io/agrestisdavid/faster-qwen3-tts-wyoming:latest https://github.com/users/agrestisdavid/packages/container/package/faster-qwen3-tts-wyoming bridge bash false --runtime=nvidia --gpus all https://raw.githubusercontent.com/agrestisdavid/faster-qwen3-tts-wyoming/main/icon.png http://[IP]:[PORT:7860]/ui/ Local Faster Qwen3-TTS with a WebUI, an OpenAI-compatible streaming API, and an integrated Wyoming TTS adapter for Home Assistant. Both interfaces share one NVIDIA GPU model instance. https://github.com/agrestisdavid/faster-qwen3-tts-wyoming/issues https://github.com/agrestisdavid/faster-qwen3-tts-wyoming https://raw.githubusercontent.com/agrestisdavid/faster-qwen3-tts-wyoming/main/templates/faster-qwen3-tts-wyoming.xml https://raw.githubusercontent.com/agrestisdavid/faster-qwen3-tts-wyoming/main/README.md AI: HomeAutomation: Qwen Qwen3 TTS Wyoming Home Assistant speech voice CUDA OpenAI Unraid 6.12 or newer, the NVIDIA Driver plugin, a CUDA-capable NVIDIA GPU, and internet access for the first model download. The default 1.7B model is recommended for quality; use the 0.6B CustomVoice model for lower VRAM usage. 1.0.0: Initial release. 2026-07-30 6.12.0 MIT