{
  "title": "audio.mp3",
  "transcript": "Welcome to the RealTest Speech Transcription Benchmark version 1.0, automated speech recognition transforms spoken audio into accurate, searchable text across podcasts, interviews, and video captions.  accuracy rate, high quality speech to text must preserve numerical precision,  terminology, and sentence structure without hallucination.",
  "duration_seconds": 39,
  "segments": [
    {
      "id": 0,
      "end": 12.96,
      "text": " Welcome to the RealTest Speech Transcription Benchmark version 1.0, automated speech recognition transforms spoken audio into accurate, searchable text across podcasts, interviews, and video captions.",
      "start": 0
    },
    {
      "id": 1,
      "end": 38.72,
      "text": " accuracy rate, high quality speech to text must preserve numerical precision,  terminology, and sentence structure without hallucination.",
      "start": 30
    }
  ],
  "translations": {},
  "insights": null
}