LogoAI Music API
Sign In

Model endpoint · POST /v2/minimax/generate

MiniMax Music API

Create full songs from a style prompt and your own lyrics with MiniMax Music 2.5 — an asynchronous REST endpoint with task polling and webhook callbacks.

30 credits per request ≈ $0.30

MiniMax Music 2.5 behind one REST call

The MiniMax Music API exposes MiniMax Music 2.5 through the same account, API key and asynchronous task model as the rest of the platform. Submit a prompt describing the track you want together with lyrics — required for MiniMax Music 2.5 at 10–3000 characters — and the endpoint returns a task_id immediately.

Results arrive through the dedicated MiniMax feed endpoint. Poll GET /v2/minimax/feed?workId=... with your task_id until the status is SUCCESS or FAILED, or pass callback_url and let the platform POST the final result to your webhook. A completed task returns the generated audio URL in the response field.

Quickstart: generate and check the feed

POST to /v2/minimax/generate, keep the returned task_id, then query the MiniMax feed until the song is ready.

curl
curl -X POST "https://udioapi.pro/api/v2/minimax/generate" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "prompt": "Cinematic orchestral pop, emotional female vocal, warm piano",
    "lyrics": "Verse 1: The city lights are fading\nChorus: We rise into the morning glow"
  }'

curl -X GET "https://udioapi.pro/api/v2/minimax/feed?workId=YOUR_TASK_ID"

Poll every 5–10 seconds with a bounded timeout. Optional bitrate and sample_rate parameters control the output encoding.

MiniMax Music 2.5 request parameters

The generate endpoint accepts a compact parameter set. One request costs 30 credits — about $0.30 at 1 USD = 100 credits.

ParameterTypeRequiredNotes
promptstringrequiredDescribes the track — genre, mood, instrumentation
lyricsstringrequired10–3000 characters; required for MiniMax Music 2.5
bitratenumberoptionalAudio bitrate, e.g. 256000
sample_ratenumberoptionalAudio sample rate, e.g. 44100
callback_urlstringoptionalPublic HTTPS webhook for the final result
publicbooleanoptionalWhether the task may be public

The response returns data.task_id and status IN_PROGRESS. Query /v2/minimax/feed?workId=<task_id> until status reaches SUCCESS (audio URLs in response) or FAILED.

What you can build

MiniMax tasks live alongside the rest of the audio API.

Lyric-accurate songs

Supply full lyrics — required at 10–3000 characters — plus a style prompt for deterministic vocal content.

Read the docs

Instrumental or vocal tracks

Steer genre, mood and arrangement through the prompt while the model handles composition.

Read the docs

More engines, one key

The same API key also calls the chirp-v6 family on /v2/generate when you need a different sound.

Read the docs

Extend generated songs

Continue eligible tracks with the extend endpoint once the source task completes.

Read the docs

WAV conversion

Convert generated audio to WAV for delivery in production pipelines.

Read the docs

Callbacks and status

Poll the MiniMax feed or deliver results to your infrastructure via callback_url.

Read the docs

How it works

From API key to a playable track in four steps.

  1. 1

    Create an API key

    Sign in, open the dashboard and copy a key. Keep it in a server-side secret — never in browser code.

  2. 2

    Submit prompt and lyrics

    POST /v2/minimax/generate with prompt, lyrics and optional encoding fields. Save data.task_id.

  3. 3

    Poll the MiniMax feed

    GET /v2/minimax/feed?workId=<task_id> every 5–10 seconds, or set callback_url for a webhook.

  4. 4

    Download the audio

    On SUCCESS the response field contains the audio URL. Persist it to your own storage.

Flat per-request pricing

Each MiniMax Music 2.5 request costs 30 credits — about $0.30 at 1 USD = 100 credits. No subscription is required for API access; top up pay-as-you-go credits when you need them.

Credits never expire. Rejected requests — for example missing or invalid lyrics — return a 400 and are not billed as a completed task.

Limits and notes

  • Generation is asynchronous; the POST returns a task_id, not audio.
  • Lyrics are required for MiniMax Music 2.5 and must be 10–3000 characters.
  • Poll /v2/minimax/feed every 5–10 seconds with a bounded timeout, or use callback_url.
  • Audio URLs are temporary — download results to your own storage.
  • MiniMax is an additional engine; the default music-generation endpoint remains /v2/generate.

udioapi.pro is an independent third-party service. It is not affiliated with, endorsed by, or operated by Suno, Udio, or MiniMax.

MiniMax Music API FAQ

Answers about parameters, billing and the asynchronous workflow.

Is this the official MiniMax API?

No. udioapi.pro is an independent third-party API service with its own endpoints, accounts, credits, terms and support. It is not affiliated with, endorsed by, or operated by MiniMax, Suno or Udio.

How much does a MiniMax Music 2.5 request cost?

Each request to /v2/minimax/generate costs 30 credits — about $0.30 at 1 USD = 100 credits.

Are lyrics required?

Yes. MiniMax Music 2.5 requires a lyrics field of 10–3000 characters. Requests with missing or invalid lyrics are rejected with a 400 response.

How do I get the result?

The generate response returns data.task_id. Poll GET /v2/minimax/feed?workId=<task_id> every 5–10 seconds until SUCCESS, or set callback_url to receive a webhook. The audio URL then appears in the response field.

What statuses will I see?

Tasks start as IN_PROGRESS, then end as SUCCESS or FAILED. A failed task reports error_message in the feed, and rejected requests are not billed as completed tasks.

Is there a cheaper model for drafts?

Yes. The chirp-v6 family on /v2/generate includes chirp-v6-mini at 10 credits per request — the fast, economical tier for prototyping.

Keep exploring