Transcribe audio | milliseconds.ai documentation

{
  "language": null,
  "segments": [
    {
      "start": 0,
      "end": 2.4,
      "text": "Hello, thanks for calling support.",
      "words": [
        {
          "word": "Hello,",
          "start": 0,
          "end": 0.72,
          "speaker": "SPEAKER_00"
        },
        {
          "word": "thanks",
          "start": 0.72,
          "end": 1.04,
          "speaker": "SPEAKER_00"
        },
        {
          "word": "for",
          "start": 1.04,
          "end": 1.2,
          "speaker": "SPEAKER_00"
        }
      ],
      "speaker": "SPEAKER_00"
    },
    {
      "start": 4,
      "end": 10.24,
      "text": "Hi, my invoice for September shows two charges of $49, and I only have one account.",
      "words": [
        {
          "word": "Hi,",
          "start": 4,
          "end": 4.56,
          "speaker": "SPEAKER_01"
        },
        {
          "word": "my",
          "start": 4.56,
          "end": 4.8,
          "speaker": "SPEAKER_01"
        }
      ],
      "speaker": "SPEAKER_01"
    }
  ],
  "duration": 19.0075,
  "text": "Hello, thanks for calling support. How can I help you today? Hi, my invoice for September shows two charges of $49, and I only have one account. I see the duplicate charge. I will refund one of them now. You should see it in three to five business days. Great. Thank you very much for your help."
}

Send one audio file as the raw request body. Any format ffmpeg decodes works (mp3, m4a, wav, ogg, opus, flac, webm). The call is synchronous and returns the whole transcript, with word timestamps and, when diarize is true, speaker labels. One call takes up to 60 minutes of audio. Price: $0.12 per audio hour with diarization,$ 0.10 without, per second rounded up. The response carries x-input-tokens (input tokens billed) and x-inference-ms.

Send one audio file as the raw request body. Any format ffmpeg decodes works (mp3, m4a, wav, ogg, opus, flac, webm). The call is synchronous and returns the whole transcript, with word timestamps and, when `diarize` is true, speaker labels. One call takes up to 60 minutes of audio. Price: $0.12 per audio hour with diarization, $0.10 without, per second rounded up. The response carries `x-input-tokens` (input tokens billed) and `x-inference-ms`.

Authentication

AuthorizationBearer

A milliseconds API key, test_sk-{namespace}-{entropy} or prod_sk-{namespace}-{entropy}. Every /v1 route requires it.

Query parameters

diarizebooleanOptionalDefaults to false

Label speakers (SPEAKER_00 to SPEAKER_03). Default false. Send true for speaker labels, at the higher price.

Request

The audio file bytes. Content type audio/* or application/octet-stream.

Response

The transcript

languageany

Always null: the model does not report a language code.

segmentslist of objects

durationdouble

Audio duration in seconds. Billing uses it.

textstring

The segment texts joined with spaces.

Errors

400

Bad Request Error

415

Unsupported Media Type Error

502

Bad Gateway Error

503

Service Unavailable Error