Skip to main content
Live transcripts allow you to receive real-time transcription data as it happens during a meeting. This feature streams transcription data to your webhook endpoint or WebSocket connection.

​
Configuration

To enable live transcripts, you need to configure the live_transcription_required field when creating a bot.

​
Webhook Configuration

{
  "live_transcription_required": {
    "webhook_url": "https://your-domain.com/transcripts"
  }
}

​
WebSocket Configuration

{
  "live_transcription_required": {
    "websocket_url": "wss://your-domain.com/transcripts"
  }
}

​
Live Audio Streaming

For live audio streaming, configure the live_audio_required field: 
{
  "live_audio_required": {
    "websocket_url": "wss://your-domain.com/audio"
  }
}

​
Webhook Payload Format

When using webhooks, your endpoint will receive POST requests with the following payload structure: 
{
  "speakerName": "string",
  "timestamp": "2024-01-15T14:30:00.000Z",
  "transcript": "string",
  "words": [
    {
      "word": "string",
      "start": "number",
      "end": "number",
      "confidence": "number",
      "speaker": "string",
      "punctuated_word": "string",
      "speech_confidence": "number"
    }
  ]
}

​
Payload Fields

FieldTypeDescription
speakerNamestringName or identifier of the speaker
timestampstringISO 8601 timestamp when the transcript was generated
transcriptstringThe transcribed text for this segment
wordsarrayArray of word-level details with timing and confidence scores

​
Word Object Fields

FieldTypeDescription
wordstringThe original word
startnumberStart time of the word in seconds
endnumberEnd time of the word in seconds
confidencenumberConfidence score for the word (0-1)
speakerstringSpeaker identifier for the word
punctuated_wordstringWord with punctuation
speech_confidencenumberSpeech confidence score for the word (0-1)

​
WebSocket Message Format

When using WebSocket connections, you’ll receive messages in the same format as the webhook payload: 
{
  "speakerName": "John Doe",
  "timestamp": "2024-01-15T14:30:00.000Z",
  "transcript": "Hello, welcome to the meeting.",
  "words": [
    {
      "word": "Hello",
      "start": 0.0,
      "end": 0.8,
      "confidence": 0.95,
      "speaker": "speaker_1",
      "punctuated_word": "Hello,",
      "speech_confidence": 0.92
    },
    {
      "word": "welcome",
      "start": 1.0,
      "end": 1.6,
      "confidence": 0.88,
      "speaker": "speaker_1",
      "punctuated_word": "welcome",
      "speech_confidence": 0.85
    }
  ]
}

​
Example Bot Creation with Live Transcripts

curl -X POST "https://api.meestream.ai/v1/bots/create" \
  -H "Authorization: Token YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "meeting_link": "https://zoom.us/j/123456789",
    "audio_required": true,
    "video_required": false,
    "bot_message": "Hello, I am the meeting assistant.",
    "live_transcription_required": {
      "webhook_url": "https://your-domain.com/transcripts"
    },
    "live_audio_required": {
      "websocket_url": "wss://your-domain.com/audio"
    }
  }'

​
Webhook Endpoint Requirements

Your webhook endpoint should:
  • Accept POST requests
  • Return a 200 status code to acknowledge receipt
  • Handle the payload format described above
  • Be publicly accessible (HTTPS required)
  • Respond within 5 seconds

​
WebSocket Connection Requirements

Your WebSocket server should:
  • Accept WSS (secure WebSocket) connections
  • Handle the message format described above
  • Maintain persistent connections
  • Implement proper error handling and reconnection logic

​
Error Handling

If your webhook endpoint is unavailable or returns an error:
  • The system will retry failed deliveries
  • Failed deliveries are logged for debugging
  • Consider implementing a fallback mechanism

​
Security Considerations

  • Use HTTPS for webhook URLs
  • Use WSS for WebSocket connections
  • Implement authentication if needed
  • Validate incoming payloads
  • Rate limit your endpoints to prevent abuse