MiniMax Music 3

About

MiniMax Music 3 is the strongest open-weights music model on the Sogni Supernet. An 8B language-model planner composes the song frame-by-frame — melody, arrangement, and vocal phrasing decided in sequence, the way a song actually unfolds — then a flow-matching diffusion transformer renders it into 44.1 kHz stereo audio.

Describe the record you want in plain producer language — genre, BPM, key, vocal character, arrangement — and hand it a lyric sheet with [verse] and [chorus] tags. It sings in English, Mandarin, Spanish, Japanese, and more, handles duets, choirs, rap flows, and a cappella stacks, and composes its own ending: the duration you set is a ceiling, not a target.

Music 3 is the premium companion to ACE-Step 1.5: the best vocal realism, lyric adherence, and song structure available in open weights, at a genuinely premium price (a song costs roughly twenty times an ACE-Step Turbo draft of the same length). For fast, cheap iterations and musician-grade controls like exact BPM and key, ACE-Step Turbo remains the workhorse.

Pricing

Use pay-as-you-go Spark packs for each track (1 Spark = $0.005), or choose a flat-rate Sogni plan for credit-free fair-use generation in the app.

API

One Sogni API key reaches every model on the Supernet — call MiniMax Music 3 with the exact model id.

import { SogniClient } from '@sogni-ai/sogni-client';

const client = await SogniClient.createInstance({
  appId: crypto.randomUUID(),
  apiKey: process.env.SOGNI_API_KEY,
  network: 'fast',
});

const project = await client.projects.create({
  type: 'audio',
  modelId: 'minimax_music3',
  positivePrompt: 'dreamy synthwave with a warm bassline',
  lyrics: undefined, // omit for instrumental
  numberOfMedia: 1,
  duration: 30,
});

const [url] = await project.waitForCompletion();
console.log(url); // result link — download within 24h

curl https://api.sogni.ai/v1/creative-agent/workflows \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $SOGNI_API_KEY" \
  -d '{
    "input": {
      "title": "First render",
      "steps": [{
        "id": "step1",
        "toolName": "generate_music",
        "arguments": { "prompt": "dreamy synthwave with a warm bassline", "model": "minimax_music3", "duration": 30 }
      }]
    },
    "confirm_cost": true
  }'

REST workflow alias: minimax_music3. Pin the exact id with the SDK. Full reference at docs.sogni.ai.