232 likesNew · 19d old
Compact 0.6B TTS model now sounds more real, 23 languages
What is Audio8-TTS-Realistic-Air?
It turns text into spoken audio with natural prosody and can clone a voice from a short reference recording. At 601 M parameters it supports 23 languages and runs on a single GPU.
Who it is for
Developers building voice assistants, localization pipelines, or any multilingual audio app can get high-quality speech without massive hardware.
How to try it
Open the interactive Playground on the Hugging Face model page.