Dramabox

A prompt-driven expressive TTS where natural-language direction controls speaker identity and performance (emotion, pauses, laughs, breaths), with optional timbre cloning from a short reference clip.