Models
Speech to text, the language model and text to speech.
| You see | It comes from |
|---|---|
| Time to final transcript | The agent's spans |
| Time to first token | The agent's spans |
| Time to first audio | The agent's spans |
| Tokens, audio seconds, characters | The agent's usage metrics |
VoiceGateway never calls a model. Prices are not shown, only usage.
Was this page helpful?