Music Generation

Music generation is a synchronous endpoint. The upstream provider (MiniMax) streams on the server side; the proxy fully consumes the stream on the server and returns the aggregated result in a single response, so clients don’t need to handle streaming.

POST https://api.smartapi.cc/api/openapi/v1/generate/music/generations

Authentication#

Authorization: Bearer sk-your-key-here
Content-Type: application/json

Request Body#

Field	Type	Required	Description
`model`	string	Yes	Model identifier, e.g. `minimax-music-2.0`.
`lyrics`	string	Yes	Lyrics. Must not be empty. Use `\n` for line breaks; `[verse]`/`[chorus]` tags are supported.
`prompt`	string	No	Style / mood description (genre, instruments, tempo, etc.).
`output_format`	string	No	Output form, e.g. `url` (playable link) or `hex` (hex-encoded audio).
`aigc_watermark`	boolean	No	Whether to add an AIGC watermark.
`audio_setting`	object	No	Audio parameters, see below.

audio_setting optional fields:

Field	Type	Description
`sample_rate`	number	Sample rate, e.g. `16000`, `24000`, `32000`, `44100`
`bitrate`	number	Bitrate, e.g. `32000`, `64000`, `128000`, `256000`
`format`	string	Audio format, e.g. `mp3`, `wav`, `pcm`

Which fields each model accepts and their allowed values can be found in the Models page; you can also select a model in the Playground and open “Custom request body / Debug” to inspect the actual request body.

Example#

curl https://api.smartapi.cc/api/openapi/v1/generate/music/generations \
  -H "Authorization: Bearer $SMARTAPI_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "minimax-music-2.0",
    "prompt": "Upbeat electronic pop with bright synths and punchy drums",
    "lyrics": "[verse]\nWe run through the night\nChasing the wind\n[chorus]\nLight up the whole sky\nNever look back",
    "output_format": "url",
    "audio_setting": { "sample_rate": 44100, "bitrate": 256000, "format": "mp3" }
  }'

Response#

The upstream raw JSON is passed through as-is. When output_format=url, it usually contains a playable link:

{
  "data": { "audio_url": "https://.../result.mp3", "status": 2 },
  "base_resp": { "status_code": 0, "status_msg": "success" }
}

When output_format=hex, the audio is returned as a hex string (e.g. data.audio); decode it into an audio file yourself. If upstream validation fails or errors out, the endpoint returns the upstream’s raw error message for easier troubleshooting.

lyrics is required; a missing value returns a parameter error immediately. Music generation can take a while, so set a sufficiently long request timeout.

Billing#

Music models are billed in credits (not tokens). minimax-music-2.0 is billed per call: a fixed number of credits per song (up to 5 minutes), independent of lyrics length or audio_setting, and only charged on success. See per-model pricing on the Models page and billing details in Billing.