Audio Transcription

Transcribe audio files with precise per-word timestamps using faster-whisper. Supports auto language detection and all common audio formats.

POST /transcribe/detect

Authentication

Bearer token in the Authorization header.

Request

Content-Type: multipart/form-data

Field	Type	Required	Description
file	File	Required	Audio file to transcribe. Accepts MP3, WAV, M4A, FLAC, OGG, and WebM.
language	string	Optional	ISO 639-1 language code (e.g. "en", "fr", "de"). Omit for auto-detection.

Response

Field	Type	Description
text	string	Full transcription text.
language	string	Detected or specified language code.
language_probability	number	Confidence of the language detection (0.0 – 1.0).
duration	number	Audio duration in seconds.
words	array	Per-word timestamps. Each entry: { word, start, end, probability }.
segments	array	Sentence-level segments. Each entry: { id, start, end, text }.

Example

Request

curl -X POST https://api.cacuda.com/transcribe/detect \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -F "file=@recording.mp3"

Response 200 OK

{
  "text": "Hello, this is a test.",
  "language": "en",
  "language_probability": 0.98,
  "duration": 3.42,
  "words": [
    {
      "word": "Hello,",
      "start": 0,
      "end": 0.42,
      "probability": 0.99
    },
    {
      "word": "this",
      "start": 0.44,
      "end": 0.62,
      "probability": 0.98
    },
    {
      "word": "is",
      "start": 0.64,
      "end": 0.72,
      "probability": 0.99
    },
    {
      "word": "a",
      "start": 0.74,
      "end": 0.8,
      "probability": 0.97
    },
    {
      "word": "test.",
      "start": 0.82,
      "end": 1.1,
      "probability": 0.99
    }
  ],
  "segments": [
    {
      "id": 0,
      "start": 0,
      "end": 1.1,
      "text": "Hello, this is a test."
    }
  ]
}

Error Codes

Code	Description
400	No file provided or unsupported audio format.
401	Missing or invalid API key.
402	Insufficient credits.
503	Transcription service unavailable.

Notes

Transcription time scales roughly linearly with audio duration.

For best results with non-English audio, specify the language parameter.

Try it in the scanner

Test this endpoint interactively with real data. No code required.

Open Scanner