Description
Voice assistants and podcasts suffer as standard TTS waits for the full clip and sounds robotic. Dia2 is an open TTS model that streams conversational audio in real time.
It plays as it generates with multi-speaker dialogue and nonverbal sounds like laughter, open-weight for local use.
Streaming:Instant audio.
Speakers:Natural dialogue.
Nonverbal:Laughs and pauses.
Open weights:Local.
It plays as it generates with multi-speaker dialogue and nonverbal sounds like laughter, open-weight for local use.
Features
Streaming:Instant audio.
Speakers:Natural dialogue.
Nonverbal:Laughs and pauses.
Open weights:Local.
