# Supertonic 3 TTS German Wyoming

Supertonic 3 TTS German Wyoming is a self-hosted ai app. Local German text-to-speech for Home Assistant using Supertonic 3, native Wyoming integration and an optimized Rust/ONNX Runtime backend. It is open source and self-hosted, installable on Unraid.

Canonical page: https://ourlocalhost.com/app/6845/supertonic-3-tts-german-wyoming

## Description

Local German text-to-speech for Home Assistant using Supertonic 3, native Wyoming integration and an optimized Rust/ONNX Runtime backend. Optimized and tested on Intel N100. Includes 10 voices, custom voice support, German text normalization and low-latency phrase streaming. First start downloads missing models and voices into the persistent folders. Existing files remain available after updates. Streaming is optional; turn it off for full-text synthesis. Audio is 44100 Hz, mono PCM16. Connect Home Assistant to the Wyoming port; HTTP provides health and voice information. Before installation, choose unused host ports and ensure both folders are writable by UID/GID 99:100.

## Platform availability

Available on 1 of 6 install platforms: Unraid.

## Health score

42/100 (Fair), based on 3 of 4 factors.
- Maintained: 72 (last commit 3 days ago)
- Popular: 0 (0 GitHub stars)
- Easy to install: 35 (1 install platform, install notes, Docker image)

## Links

- Unraid: https://ca.unraid.net/apps/supertonic3-tts-german-wyoming-05cuqsm0okej76?q=supertonic3-tts-german-wyoming
- GitHub: https://github.com/marco-taylor/supertonic3-tts-german-wyoming

## Common questions

### How do I connect this service in Home Assistant?

Add the Wyoming Protocol integration using the Docker host's IP address or hostname and port 10200, then choose Supertonic as the TTS service in the voice assistant pipeline. HTTP port 8881 is only for health and voice discovery and is not required by the Wyoming integration.

### What permissions do the model and voice folders need for bind mounts?

Grant the container's UID/GID read and write access to both persistent folders, or run the container with a UID/GID matching the directory owner. The included Unraid template uses UID/GID 99:100, so both folders must be writable by that user and group.

### Does the container download models on first start, and what happens after updates?

On first start it downloads approximately 380 MiB of missing model and voice assets from the official Supertonic 3 revision. Existing files are reused after updates, and only missing files are downloaded.

### How are custom voices handled?

Custom Supertonic 3 voice JSON files are detected automatically from the persistent voice directory.

### Is phrase streaming mandatory?

No, low-latency phrase streaming is optional. The TTS_STREAMING environment variable is exposed in the example configuration to enable or disable it.

---
Source: Our Local Host (https://ourlocalhost.com/app/6845/supertonic-3-tts-german-wyoming), the self-hosted and open-source app directory.