Muyan-TTS is a trainable TTS model for podcast applications, pre-trained on over 100,000 hours of audio data, enabling zero-shot synthesis and speaker adaptation with high-quality voice generation.
