JSONKit
AI & LLM

Speech-to-Text Transcription (Whisper) JSON Example

A JSON example of a speech-to-text transcription response — includes the full text, detected language, duration, and timestamped segments. Copy-ready for Whisper and audio transcription APIs.

{
  "task": "transcribe",
  "language": "english",
  "duration": 8.47,
  "text": "Welcome to JSONKit. This is a quick demo of audio transcription.",
  "segments": [
    {
      "id": 0,
      "start": 0,
      "end": 2.4,
      "text": " Welcome to JSONKit.",
      "avg_logprob": -0.21
    },
    {
      "id": 1,
      "start": 2.4,
      "end": 8.47,
      "text": " This is a quick demo of audio transcription.",
      "avg_logprob": -0.18
    }
  ]
}

Field Reference

languagerequiredstringDetected (or specified) spoken language
durationrequirednumberAudio length in seconds
textrequiredstringThe full transcript as a single string
segmentsoptionalarray<object>Timestamped chunks with start/end seconds — present in verbose_json format
segments[].avg_logproboptionalnumberAverage log-probability; very low values flag low-confidence audio

Variants

Word-level timestampsRequest granular word timings for captions and karaoke-style highlighting.
Word-level timestamps
{
  "text": "Welcome to JSONKit.",
  "words": [
    {
      "word": "Welcome",
      "start": 0,
      "end": 0.6
    },
    {
      "word": "to",
      "start": 0.6,
      "end": 0.8
    },
    {
      "word": "JSONKit",
      "start": 0.8,
      "end": 1.4
    }
  ]
}

Common Use Cases

  • →Generating subtitles and captions from audio or video
  • →Indexing podcasts and meetings for full-text search
  • →Building voice interfaces that transcribe then feed an LLM
whisperspeech to texttranscriptionsttaudioai

Validate or format this JSON

One click loads this exact example into the tool — no copy-paste needed. Format it, validate it, explore the tree, or generate TypeScript types instantly.

FormatValidateExplore treeGenerate TypeScriptMinify

Frequently Asked Questions

The 'json' response returns just the transcript text. The 'verbose_json' format adds language, duration, and the segments array with per-segment timestamps and confidence — use it whenever you need timing for captions or search.

Request timestamp_granularities including 'word' (and/or 'segment'). The response then includes a words array with start/end seconds per word, which is ideal for precise captions and audio highlighting.

Yes. Whisper auto-detects the language and can transcribe many languages; it can also translate speech to English with the translate task. Set the language explicitly if you know it to improve accuracy and speed.

Related JSON Examples

AI & LLMAI Image Generation Request & ResponseText-to-image generation request and response with prompt, size, and image URLs.AI & LLMOpenAI Chat Completion ResponseOpenAI Chat Completions API response with choices, message, finish_reason, and usage.IoT & SensorsIoT Device TelemetrySensor readings from an IoT device with GPS, battery, and environmental metrics.