TADA - Text-Acoustic Dual Alignment LLM

A demo of tada-3b-ml — a text-to-speech model that clones voice, emotion, and timing from a short audio prompt.

How to use: Choose a voice prompt (or upload your own), enter text, and click Generate. The model will encode the prompt and generate speech in one step.

Voice Prompt

Pick a preset or upload / record your own

Transcript

Pick a preset or type your own below