Zero-shot voice cloning across 24 languages and 21 Chinese dialects, plus instruction-driven voice design and semantic and acoustic speech editing.
No check-ins yet.