125 likesNew · 17d old
Tiny 0.6B TTS model now sounds real in 23 languages
What is Audio8-TTS-Realistic-0.6b?
Audio8-TTS-Realistic-0.6B turns text into natural-sounding speech and can clone a voice from a short reference clip. It runs on a 601 million-parameter model plus a separate audio codec, supporting zero-shot generation in 23 languages.
Who it is for
Developers building voice assistants, language-learning apps, or any multilingual audio content can get realistic speech without massive models.
How to try it
Open the interactive Playground on the Hugging Face repo page.