Audio transcriptions
POST /api/v1/audio/transcriptions: turn speech into text, with segments, timestamps and subtitles. For examples and explanations, see the Speech to text guide.
Create transcription
POST /api/v1/audio/transcriptions
Transcribes an audio file. The body is multipart/form-data.
Form fields
filefilerequired
Audio in any common format (MP3, WAV, OGG, FLAC, M4A, …).
modelstringrequired
arklabsai/whisper-large-v3-turbo.languagestringoptional
ISO-639-1 code of the spoken language, such as pl or en.
response_formatstringoptional
json (default), text, verbose_json, srt or vtt.temperaturenumber, 0–1optional
Default 0.
timestamp_granularitiesarrayaccepted, no effect
verbose_json always returns segment timestamps; word timestamps are not available.Returns {"text", "usage"} for json; the text plus task, language, duration and segments (each with id, start, end, text) for verbose_json; plain text for text, srt and vtt. Billed by audio duration.
Request
curl https://api.ark-labs.cloud/api/v1/audio/transcriptions \
-H "Authorization: Bearer $ARK_API_KEY" \
-F model=arklabsai/whisper-large-v3-turbo \
-F file=@call.mp3 \
-F language=en \
-F response_format=verbose_jsonResponse
{
"task": "transcribe",
"language": "english",
"duration": 12.48,
"text": "Thanks for calling. How can I help?",
"segments": [
{"id": 0, "seek": 0, "start": 0.0, "end": 3.1, "text": "Thanks for calling.", …},
{"id": 1, "seek": 310, "start": 3.4, "end": 5.2, "text": "How can I help?", …}
],
"usage": {"type": "duration", "seconds": 12.48, "credits_used": 50, "credit_balance": 4999942}
}