Skip to content

MiniMax Music โ€‹

Updated

Endpoint: POST https://vod.intl.tencentcloudapi.comAction: CreateAigcAudioTask

MiniMax Music is a music generation model that creates songs from a prompt and supplied lyrics, outputting MP3 (44.1 kHz). Versions 3.0, 2.6, 2.5, and 2.0 are available.

Basic Information โ€‹

ItemValue
ModelNameMiniMaxMusic
ModelVersion3.0 / 2.6 / 2.5 / 2.0
SceneTypemusic
Output bitrateCheck in the metadata of the generated file
Output lengthVaries by model and request options; check in the result metadata
Output specMP3, 44.1kHz stereo
Task completionConfirmed by the completion status of the query API

You can supply lyrics directly.

Request Parameters โ€‹

ParameterRequiredTypeDescription
ModelNameRequiredStringFixed value MiniMaxMusic
ModelVersionOptionalString3.0 / 2.6 / 2.5 / 2.0
SceneTypeRequiredStringmusic
PromptRequiredStringPrompt for the song's mood and style
AdditionalParametersOptionalStringlyrics (lyrics), instrumental (whether to remove vocals, Boolean)

Request Examples โ€‹

json
{
  "SubAppId": 123456789,
  "ModelName": "MiniMaxMusic",
  "ModelVersion": "2.6",
  "SceneType": "music",
  "Prompt": "a cheerful pop song, upbeat tempo",
  "AdditionalParameters": "{\"lyrics\": \"Sunlight on the windowsill, a new day has begun\"}"
}
json
{
  "SubAppId": 123456789,
  "ModelName": "MiniMaxMusic",
  "ModelVersion": "3.0",
  "SceneType": "music",
  "Prompt": "lo-fi hip hop beat for study background",
  "AdditionalParameters": "{\"instrumental\": true}"
}

Response Example โ€‹

json
{
  "AigcAudioTask": {
    "Status": "FINISH",
    "ErrCode": 0,
    "Output": {
      "AudioInfos": [
        {
          "FileUrl": "http://<host>.vod2.myqcloud.com/.../aigcAudioGenFile.mp3"
        }
      ]
    }
  }
}

Model-Specific Audio Options โ€‹

ModelFieldDescription
MiniMaxMusiclyricsLyrics
MiniMaxMusicinstrumentalWhether to remove vocals. Boolean
MiniMaxMusicsample_rate / bitrateOutput audio spec

Special Settings โ€‹

Version Selection โ€‹

ModelVersionCharacteristics
3.03.0 series
2.62.6 series
2.5Older version
2.0Initial version

Lyrics Specification (lyrics) โ€‹

Lyrics are passed through AdditionalParameters's lyrics field. In Prompt, write the song's mood and style.

json
{
  "SubAppId": 123456789,
  "ModelName": "MiniMaxMusic",
  "ModelVersion": "3.0",
  "SceneType": "music",
  "Prompt": "a bright and cheerful pop song",
  "AdditionalParameters": "{\"lyrics\": \"Sunlight on the windowsill, a new day has begun\"}",
  "OutputConfig": {
    "StorageMode": "Temporary",
    "OutputAudioFormat": "mp3"
  }
}

AdditionalParameters โ€‹

FieldTypeDescription
lyricsStringLyrics
instrumentalBooleanRemove vocals. Generate accompaniment only
sample_rateIntegerOutput sample rate
bitrateIntegerOutput bitrate

The output bitrate is fixed at 32 kbps

MiniMaxMusic outputs 32 kbps MP3. For 192 kbps output, use GL.

Instrumental Generation โ€‹

Set instrumental to true.

json
{
  "SubAppId": 123456789,
  "ModelName": "MiniMaxMusic",
  "ModelVersion": "3.0",
  "SceneType": "music",
  "Prompt": "lo-fi hip hop beat for study background",
  "AdditionalParameters": "{\"instrumental\": true}"
}

Output Format โ€‹

Use OutputConfig.OutputAudioFormat to specify the result file format.

json
{
  "OutputConfig": {
    "StorageMode": "Temporary",
    "OutputAudioFormat": "mp3"
  }
}

Additional Options โ€‹

VOD's vocal control uses instrumental, and MPS uses a is_instrumental Boolean value. MiniMaxMusic's output is 32 kbps MP3.

Built with VitePress ยท Deployed on Tencent EdgeOne Pages