ChatTTS Colab
A ChatTTS-based text-to-speech app that runs locally on Windows and macOS or in Google Colab, with saved voices, character reading and streaming playback.
Alternative to:
ChatTTS Colab turns text into speech for people producing narration or scripts with different character voices. It builds on ChatTTS and provides a browser interface for comparing voices and generating long audio passages.
- Voice selection: Generate batches of voice candidates and save the ones you want to reuse; the related ChatTTS_Speaker library supports searches by gender, age and voice characteristics.
- Character reading: Assign different voices to characters in a script, with language-model-assisted script generation also available.
- Long audio and streaming: Generate extended spoken content and start listening while speech generation continues.
- Text preparation: Basic processing handles numbers and punctuation that can cause reading errors.
The app runs locally on macOS and has an offline Windows package. Local speech generation loads the ChatTTS model on your machine. Standard local use downloads model files from Hugging Face, while the offline package includes model files that the app can load locally.
Google Colab is a separate cloud execution option accessed through the browser; speech generation there runs on Colab rather than your computer.
