Transcripts

​

Get Transcript History

​

Endpoint

GET https://api.mentra.glass/api/transcripts/:appSessionId

GET https://devapi.mentra.glass/api/transcripts/:appSessionId

GET http://localhost:8002/api/transcripts/:appSessionId

​

Headers

X-API-Key: <appApiKey>
X-Package-Name: <packageName>

​

Parameters

​

Query Parameters

​

Response

{
  "language": "en-US",
  "segments": [
    {
      "speakerId": "speaker-1",
      "resultId": "1705748400000-abc123",
      "text": "Hello, how are you today?",
      "timestamp": "2024-01-20T10:00:00.000Z",
      "isFinal": true
    },
    {
      "speakerId": "speaker-1",
      "resultId": "1705748403000-def456",
      "text": "I'm doing great, thanks!",
      "timestamp": "2024-01-20T10:00:03.000Z",
      "isFinal": true
    }
  ]
}

{
  "error": "duration, startTime, or endTime is required"
}

{
  "error": "Session not found"
}

{
  "error": "Error fetching transcripts"
}

​

Implementation

• File: packages/cloud/src/routes/transcripts.routes.ts:18-70
• Service: Uses TranscriptionManager.getTranscriptHistory()
• Session: Extracts user session ID from app session ID by splitting on ’-‘

​

Time Range Options

1. Duration-based (most common):
   GET /api/transcripts/session-123?duration=300&language=en-US
   
   Returns last 5 minutes (300 seconds) of transcripts
2. Time range:
   GET /api/transcripts/session-123?startTime=2024-01-20T10:00:00Z&endTime=2024-01-20T10:30:00Z
   
   Returns transcripts within specific time range

​

Transcript Segment Structure

• speakerId: Optional speaker identifier
• resultId: Unique identifier for this transcript segment
• text: The transcribed text
• timestamp: ISO timestamp when the segment was created
• isFinal: Whether this is a final or interim result

​

Language Support

• en-US - English (US) - default
• es-ES - Spanish (Spain)
• fr-FR - French (France)
• de-DE - German (Germany)
• it-IT - Italian (Italy)
• pt-BR - Portuguese (Brazil)
• ja-JP - Japanese (Japan)
• ko-KR - Korean (Korea)
• zh-CN - Chinese (Simplified)

​

Session ID Format

​

Error Codes

​

Notes

• Transcripts are stored in memory per user session
• Historical data availability depends on session duration
• Language filtering returns only transcripts in the specified language
• Apps should authenticate using API key headers (though validation is not shown in the current implementation)
• Console logs are present for debugging (lines 26, 59)