voice-to-voice
Speech models that take audio input and generate audio output directly, preserving tone and cadence, rather than transcribing to text and then synthesizing speech.
As defined in Jesse Zhang - Building Decagon - [Invest Like the Best, EP.443]
Where it comes up · 1
Most important claims · 2 of 2Search all claims for “voice-to-voice” →
- Voice-to-voice is more powerful than text-to-speech because it takes in the whole audio of what was said, including cadence and how upset the speaker is, and it reduces latency compared with routing through text.Jesse Zhang - Building Decagon - [Invest Like the Best, EP.443]
- Voice-to-voice models currently have a hallucination rate roughly 8x higher than text because voice generates far more tokens per sentence, and more tokens create more opportunities for errors, which blocks direct enterprise use.Jesse Zhang - Building Decagon - [Invest Like the Best, EP.443]