tkinterPiperTTS

A Windows application wrapper for PiperTTS with auto-download and installation of models from Hugging Face. Manual model addition, text-to-speech playback, config adjustment, and WAV export.

Add custom models to models/ (in the app folder). folder within root.

Get voice models manually

• Piper (official) voice list
• Hugging Face: rhasspy/piper-voices

App Usage

• Download:
  • Single Executable (Not Recommended, Slow Startup)
  • Extracted Executable (Recommended, Instant Startup)
• Run the PiperTTS.exe.
• Select/download/add model.
• Enter text in box.
• Adjust config (volume, speed, noise, noise_w).
• Click "▶" for audio or "💾" to export as WAV.

Development Setup

• Install Python 3.10+.
• Install dependencies: pip install customtkinter piper-tts sounddevice numpy
• Clone repo and run main.py.

Functions

• load_model(): Loads selected ONNX model.
• download_voice(): Downloads model from Hugging Face.
• add_model(): Adds local models to /models.
• play(): Synthesizes and plays audio in a thread.
• export_wav(): Saves synthesized audio to WAV.
• refresh_models(): Updates model list.

WindowsPiperTTS/
├─ main.py
├─ icon.ico
├─ models/                             # create this
│  ├─ en_US-lessac-medium.onnx
│  ├─ en_US-lessac-medium.onnx.json
│  ├─ en_GB-cori-high.onnx
│  └─ en_GB-cori-high.onnx.json
└─ README.md

Get voice-models automatically from hugginface.co

Manually upload voice models

Play or output to file