Voices, speech and transcription

Use these routes to pick a voice, turn text into audio, clone or design a voice, and turn a short recording into text.

A voice works anywhere a voice_id is accepted: spoken sends, voice broadcasts, AI agents and text-to-speech. To send a spoken message, you don't need to synthesize it first. Pass tts_body and voice_id on the send, as described in Sending basics.

Routes

Pick a voice

The list holds your cloned and designed voices, plus the platform catalog.

curl "https://api-v2.dropcowboy.com/voice/public/voices?include_urls=true" \
  -H "x-key: $DC_KEY" -H "x-secret: $DC_SECRET"

{
  "data": {
    "voices": [
      {
        "voice_id": "7d2a9e4b-1c6f-4b3a-8e5d-2f9c7a1b4e63",
        "team_id": "3f6c2a1e-8b4d-4c7a-9e2f-5a1b3c4d6e7f",
        "name": "Jordan (cloned)",
        "type": null,
        "gender": "female",
        "status": "ready",
        "pro_voice": false,
        "url": "https://media.example.com/samples/jordan.wav",
        "created_at": 1774041600000,
        "failed_at": null
      }
    ],
    "total": 1
  },
  "meta": { "request_id": "c2a6e8f4-3b9d-4c1e-8a7f-5d3b1e9c6a24" }
}

Catalog voices have a null team_id. Designed voices have type: "designed". data.total counts your account's voices, not the catalog. Timestamps are epoch milliseconds.

Only voices with status: "ready" can speak, catalog voices included:

Text-to-speech

Turn text into an audio file with any voice the list shows as ready. Use it to preview a script, to cache audio you play many times, or to produce files for another system.

curl https://api-v2.dropcowboy.com/voice/public/tts/synthesize \
  -H "x-key: $DC_KEY" -H "x-secret: $DC_SECRET" \
  -H "content-type: application/json" \
  -d '{
    "voice_id": "7d2a9e4b-1c6f-4b3a-8e5d-2f9c7a1b4e63",
    "text": "Thanks for calling Example Dental. How can I help?"
  }'

{
  "data": {
    "audio_url": "https://media.example.com/tts/1b7e3c9a.mp3",
    "expires_at": 1774045200000,
    "tts_characters": 50,
    "content_type": "audio/mpeg"
  },
  "meta": { "request_id": "c2a6e8f4-3b9d-4c1e-8a7f-5d3b1e9c6a24" }
}

The call is synchronous. audio_url works for about an hour, until expires_at, so download the file if you need it later. You're billed per character, and tts_characters reports the billed count.

Clone a voice

Get consent from the person whose voice you clone. Send one of sample_url, media_id or voice_id.

curl https://api-v2.dropcowboy.com/voice/public/voices/clone \
  -H "x-key: $DC_KEY" -H "x-secret: $DC_SECRET" \
  -H "content-type: application/json" \
  -d '{
    "name": "Jordan",
    "sample_url": "https://media.example.com/samples/jordan.wav",
    "gender": "female"
  }'

The route answers 202 and builds the clone in the background:

{
  "data": {
    "voice_id": "7d2a9e4b-1c6f-4b3a-8e5d-2f9c7a1b4e63",
    "long_job_id": "1d5b9f3e-7a2c-4e8d-b6f1-3c7a9e5d2b84"
  },
  "meta": { "request_id": "c2a6e8f4-3b9d-4c1e-8a7f-5d3b1e9c6a24" }
}

If the voice slot still needs buying, the response also carries pending_payment: true. Complete checkout in the dashboard.

Wait for it to finish

Poll GET /voice/public/voices every few seconds and find your voice_id. Stop when its status is ready or failed.

Design a voice

Describe a voice in words and hear a preview. Previews are free.

curl https://api-v2.dropcowboy.com/voice/public/voices/design \
  -H "x-key: $DC_KEY" -H "x-secret: $DC_SECRET" \
  -H "content-type: application/json" \
  -d '{
    "instructions": "Warm, friendly receptionist in her thirties, relaxed pace.",
    "text": "Thanks for calling Example Dental. How can I help?"
  }'

The route answers 202 with a long_job_id:

{
  "data": { "long_job_id": "8a4c2e6f-9b1d-4f3a-a7e5-2c8b6d4f1e90" },
  "meta": { "request_id": "c2a6e8f4-3b9d-4c1e-8a7f-5d3b1e9c6a24" }
}

Listen to the preview and save it in the dashboard. A saved design appears in GET /voice/public/voices with type: "designed".

Delete a voice

Wait until a clone's status is ready or failed before you delete it.

curl -X DELETE https://api-v2.dropcowboy.com/voice/public/voices/7d2a9e4b-1c6f-4b3a-8e5d-2f9c7a1b4e63 \
  -H "x-key: $DC_KEY" -H "x-secret: $DC_SECRET"

{
  "data": {
    "voice_id": "7d2a9e4b-1c6f-4b3a-8e5d-2f9c7a1b4e63",
    "deleted": true
  },
  "meta": { "request_id": "e4b8a2c6-9d3f-4a1e-b7c5-8f2d6a9e3b17" }
}

The voice leaves the list and can't be used in sends or synthesis. If it counted toward your monthly saved-voice subscription, that count drops by one. A voice in pending_payment is removed from your checkout cart instead. You can delete only your own voices, not catalog voices.

Transcription

Turn a short recording into text, for example a voicemail reply or a recorded message you want to check before a campaign. Send media_id or url.

curl https://api-v2.dropcowboy.com/voice/public/asr/transcribe \
  -H "x-key: $DC_KEY" -H "x-secret: $DC_SECRET" \
  -H "content-type: application/json" \
  -d '{ "media_id": "1b7e3c9a-4d2f-4a8b-9e6c-7f2a1d5b3c80" }'

{
  "data": {
    "text": "Hi, this is Jordan returning your call about the appointment.",
    "language": "en"
  },
  "meta": { "request_id": "c2a6e8f4-3b9d-4c1e-8a7f-5d3b1e9c6a24" }
}

The call is synchronous and times out after 29 seconds. Keep recordings to a few minutes and under 10 MB.

Audio formats

• media_id takes any MP3 or WAV you uploaded as media. An MP3 needs a few seconds after the upload completes before it can be transcribed. Until then it answers 409: wait two or three seconds and retry.
• url must point at a WAV file with PCM 16-bit samples. Any other format answers 400. For MP3 or other formats, upload the file as media and send media_id instead.

Errors

The errors specific to these routes are below. For the error format and everything else, see Responses, errors and limits.

Text-to-speech errors

Clone, design and delete errors

Transcription errors