ZONOS
Suite of audio models - ...
ZONOS2
ZONOS2 specializes in high quality emotive voice generation with a focus on lower latency, increased multilingual accuracy, and more faithful voice cloning.
PREVIOUS MODELS
ZONOS0.1
MODALITIES
Text-to-Speech (TTS)
ARCHITECTURE
7.7B-870M Parameters Mixture-of-Experts (MoE)
FEATURES
Zero-Shot Voice Cloning
Clones voices from short voice samples.
Highly Expressive
Realistic and emotional generations.
Multilingual Support
Support for English and Japanese.
Controllable
Fine grained control over speaker and audio characteristics.