Inworld has introduced two updates for Realtime TTS-2: Reasoning first generates delivery guidance based on the text and context before synthesizing speech, while the model update reduces generation time and improves voice consistency.
Source:Inworld AI · Read original
Leave a Reply