Audio8-TTS Released: A 0.6B Open-Source Speech Model with Zero-Shot Cloning

bigblueboo · x · 2026-07-30

Developer @SamuelZengML announced the open-source release of Audio8-TTS Preview. Despite being a compact 0.6B parameter model, it boasts impressive capabilities: supporting 11 languages, zero-shot voice cloning, and a bundled 44.1 kHz codec under an Apache 2.0 license.

Tech reviewer @bigblueboo expressed excitement to test the model in their personal Pocket-clone app, seeing if it can dethrone their current favorite, voxcpm2.

Related event: Audio8-TTS Tops Hugging Face Trending Chart(2 posts)→

Original post →

More from Models

Models channel →