piper-plus Demo

High-quality multilingual text-to-speech synthesis supporting Japanese, English, Chinese, Spanish, French, and Portuguese.

This demo uses a single multilingual ONNX model for fast CPU inference.

Select Model

Advanced Settings

0.5 2
0 1
0 1

Tips:

  • Select the language matching your input text
  • All modes use the same multilingual ONNX model
  • Adjust speed for faster/slower speech
  • Higher expressiveness = more natural variation
  • Chinese, Spanish, French, Portuguese require piper_train installed
Examples