How i get the assistant's output generated by the language model rather than the one from spoken

Based on the documentation, you can access the language model's output before it's converted to speech through the

message

event in Vapi.

javascript
vapiInstance.on('message', (message) => {
  console.log(message);
});

According to [docs.vapi.ai](https://docs.vapi.ai/examples/voice-widget), this event carries function calls and transcripts. Are you trying to access this output in a web application, server-side application, or through another specific Vapi integration? This would help me provide more targeted instructions for your use case. Sources: - [Vapi Documentation](https://docs.vapi.ai/examples/voice-widget) - [Vapi Documentation](https://docs.vapi.ai/quickstart) - [Vapi Documentation](https://docs.vapi.ai/how-vapi-works)

Hey, the Assistant output for the next response is generated by the large language model and it's triggered on the user input. I want to know, "What's confusing you here?"