# Phonon-TTS

Phonon-TTS is a self-hosted ai app. Phonon TTS runs Gradium's Phonon text-to-speech models and serves an OpenAI-compatible speech API (POST /v1/audio/speech), so any client with an "OpenAI TTS" setting and a custom base URL works unchanged. It is open source and self-hosted, installable on Unraid.

Canonical page: https://ourlocalhost.com/app/7394/phonon-tts

## Description

Phonon TTS runs Gradium's Phonon text-to-speech models and serves an OpenAI-compatible speech API (POST /v1/audio/speech), so any client with an "OpenAI TTS" setting and a custom base URL works unchanged. It can also serve the Wyoming protocol, which Home Assistant uses for local voice. It runs entirely on the CPU and is fast: Gradium's models speak faster than real time even on a Raspberry Pi. Language: PTTS_LANG is required and picks how numbers and symbols are spelled out, and which catalog voices are served; it must match the language of the checkpoint (en by default; fr for a French model, and so on). First start: downloads the default model and the voices for its language into the cache folder, and logs little until they load; watch the log. The cache folder keeps them across restarts and updates, so later starts go straight to serving. Home Assistant: set PTTS_WYOMING_ADDR to

## Platform availability

Available on 1 of 9 platforms: Unraid.

## Health score

67/100 (Good), based on 3 of 4 factors.
- Maintained: 100 (last commit 0 days ago)
- Popular: 52 (392 GitHub stars)
- Easy to install: 20 (1 install platform, Docker image)

## Links

- Unraid: https://ca.unraid.net/apps/app_1617c55141a8b6b7
- GitHub: https://github.com/gradium-ai/xn-ptts

---
Source: Our Local Host (https://ourlocalhost.com/app/7394/phonon-tts), the self-hosted and open-source app directory.