Voice Cloning

Voice cloning lives in the MiniMax route family (/v1m). Upload a sample, keep the returned voice_id, then reuse that id in MiniMax TTS requests.

Create Clone

http

POST https://api.cheapaiapi.com/v1m/voice/clone

Send this request as multipart/form-data.

filefilerequiredSource audio sample.

voice_namestringDisplay name for the cloned voice.

preview_textstringPreview text for the cloning flow.

language_tagstringLanguage label such as English.

need_noise_reductionbooleanEnable cleanup for noisy source audio.

gender_tagstringOptional tag such as male or female.

bash

curl -X POST "https://api.cheapaiapi.com/v1m/voice/clone" \
  -H "Authorization: Bearer sk_your_api_key" \
  -F 'file=@voice_sample.mp3' \
  -F 'voice_name=My Custom Voice' \
  -F 'language_tag=English' \
  -F 'preview_text=Hello world'

Create Clone Response

json

{
  "success": true,
  "cloned_voice_id": 12345
}

List Clones

http

GET https://api.cheapaiapi.com/v1m/voice/clone

json

{
  "success": true,
  "data": [
    {
      "voice_id": "12345",
      "voice_name": "My Custom Voice",
      "tag_list": ["English", "Clone"],
      "sample_audio": "https://..."
    }
  ]
}

Delete Clone

http

DELETE https://api.cheapaiapi.com/v1m/voice/clone/{voice_clone_id}

How It Connects To TTS

Once a clone exists, use its voice_id inside the MiniMax TTS route:

json

{
  "model": "speech-2.6-hd",
  "voice_setting": {
    "voice_id": "12345",
    "vol": 1,
    "pitch": 0,
    "speed": 1
  }
}

Tips For Better UX

Show upload quality guidance before the user submits the source audio.
Let the user name the clone during creation so it is reusable later.
After clone creation, offer a direct jump into MiniMax TTS with the new voice_id prefilled.
Keep list and delete actions close together so teams can manage test voices cleanly.