Skip to main content

​
Overview

The voice module provides real-time voice transcription via AssemblyAI streaming API. Users can speak into their microphone and get live text transcription in the chat interface.

​
Exports

​
thepopebot/voice

React hook for voice input: 
import { useVoiceInput } from 'thepopebot/voice';
useVoiceInput() Custom React hook that provides voice transcription state and controls. Returns:
  • isRecording (boolean) - Whether recording is active
  • transcript (string) - Current transcription text
  • startRecording() - Start voice capture
  • stopRecording() - Stop voice capture
  • error (string | null) - Error message if transcription fails

​
thepopebot/voice/actions

Server actions for voice token management: 
import { getVoiceToken } from 'thepopebot/voice/actions';
getVoiceToken() Server action that generates a temporary AssemblyAI streaming token. Returns:
  • { token: string } on success
  • { error: string } if voice transcription is not configured or token fetch fails
Requires ASSEMBLYAI_API_KEY environment variable. Token is valid for 60 seconds and is only needed at WebSocket handshake time.

​
Setup

Enable voice input by setting the AssemblyAI API key: 
# In your .env file
ASSEMBLYAI_API_KEY=your_assemblyai_api_key_here
Get your API key from AssemblyAI.

​
Usage Example

import { useVoiceInput } from 'thepopebot/voice';

export function VoiceButton() {
  const { isRecording, transcript, startRecording, stopRecording, error } = useVoiceInput();

  return (
    <div>
      <button onClick={isRecording ? stopRecording : startRecording}>
        {isRecording ? 'Stop' : 'Start'} Recording
      </button>
      {transcript && <p>Transcript: {transcript}</p>}
      {error && <p>Error: {error}</p>}
    </div>
  );
}

Build docs developers (and LLMs) love

Get started for freeTalk to us