Transcribe an audio file or URI | Makimoto

curl -X POST https://https//https/transcribe \
     -H "apikey: {{apikey}}" \
     -H "Content-Type: application/json" \
     -d '{
  "uri": "s3://your-bucket/recording.wav",
  "call_uuid": "call-abc123",
  "callback_url": "https://https//hooks.example.com/callback"
}'

{
  "data": {
    "id": "29051cac-89b5-49ec-a932-006c8cc57-0febd3.wav",
    "metadata": {
      "lang": "en",
      "mode": "POST",
      "codec": "",
      "timestamp": 1780635349,
      "sample_rate": 16000,
      "audio_duration": 571.66,
      "num_of_channels": 2,
      "transcribed_mode": "POST"
    },
    "transcript": [
      {
        "text": "Hello, how can I help you today?",
        "time_end": 2.5,
        "speaker_id": 0,
        "time_start": 0,
        "speaker_alias": "Agent"
      },
      {
        "text": "I'd like to check my account balance.",
        "time_end": 5.1,
        "speaker_id": 1,
        "time_start": 1,
        "speaker_alias": "User"
      }
    ]
  },
  "status": "S001",
  "call_uuid": "call-abc123"
}

Submit a file upload OR a URI for batched, server-side transcription. Returns the full transcript inline when callback_url is omitted; otherwise returns an accepted body and POSTs the transcript to the callback when ready.

Audio is processed through VAD + diarization + denoise locally, then shipped to the stt-model Whisper inference container.

Request

This endpoint expects an object.

uristringRequired

call_uuidstringRequired

callback_urlstringRequiredformat: "uri"

Response

Inline mode (no callback_url)

dataobject

statusstring

call_uuidstring

Errors

400

Bad Request Error

401

Unauthorized Error

413

Content Too Large Error

500

Internal Server Error

502

Bad Gateway Error

503

Service Unavailable Error

504

Gateway Timeout Error

Headers

Request

Response

Errors