F5-TTS

Flow-matching zero-shot voice cloning TTS that fakes fluent speech from a short reference clip.

F5-TTS is a non-autoregressive text-to-speech model based on flow matching with diffusion transformers, producing natural, fluent speech and zero-shot voice cloning from a few seconds of reference audio. It is notable for strong quality and speed without phoneme alignment or a separate duration model, and supports multilingual and code-switched synthesis.