EmotiVoice: Free Open Source TTS With Emotion Control

The story

What the video said

EmotiVoice, an open source speech tool, turns plain text into spoken audio with feeling. You type a line, add a mood word, and it reads the line out loud in that mood.

You want to give your app a voice that sounds happy or sad on demand, but most speech tools sound flat and cost money for every clip. You need a free one you can run yourself.

EmotiVoice is an open source text to speech engine from NetEase Youdao. It speaks Chinese and English with over two thousand voices, and it runs on your own machine with Python and PyTorch.

You give it a short style prompt like happy or sad, and the engine shifts pitch, speed, and energy to match. The repo shows demo clips and a Streamlit page where you can try it.

So EmotiVoice gives you a free, self hosted way to make speech with emotion in two languages. You can start with the Docker image or the Streamlit demo and hear the difference.

This page was generated from the Ming Dao AI publication record.