# MiniMax Music 2.5 - Core.Today AI API

> Generate full-length songs (up to ~5 minutes) with vocals, lyrics, and rich instrumentation. 14 structure tags, style-aware mixing, and an expanded instrument library including orchestral and traditional instruments.

- **Provider**: MiniMax
- **Model ID**: minimax/music-2.5
- **Category**: Audio & TTS
- **Credits**: 350 per generation (1 audio file)
- **Speed**: Medium
- **Quality**: High

## Features
- Full-length songs up to ~5 minutes (most 2:30-4:30)
- 14 structure tags: [Intro], [Verse], [Chorus], [Hook], [Drop], [Bridge]...
- Realistic vocals with breathing and pitch transitions
- Expanded instrument library incl. orchestral and traditional instruments
- Style-aware mixing (rock distortion, jazz warmth, electronic transients)
- Up to 44.1kHz / 256kbps / WAV output

## Use Cases
- Original song production from lyrics
- Background music for videos and podcasts
- Demo tracks for songwriters
- Personalized songs for events
- Social media content music

## API Endpoint

Base URL: https://api.core.today/v1
Create Prediction: POST /predictions
Get Status: GET /predictions/{job_id}

## Authentication
Header: X-API-Key: YOUR_API_KEY

## Input Parameters

### Required
- **lyrics**: string - Lyrics (1-3,500 chars). Use \n for line breaks, \n\n for pauses. Structure tags: [Intro], [Verse], [Pre Chorus], [Chorus], [Post Chorus], [Hook], [Drop], [Bridge], [Solo], [Inst], [Interlude], [Break], [Transition], [Outro]

### Optional
- **prompt**: string - Style description (max 2,000 chars): genre, mood, vocal description, BPM, instruments, era reference, production style. Be specific — '1980s Minneapolis sound' beats 'retro'
- **sample_rate**: integer (default: 44100) - Sample rate for the generated music
  Options: 16000, 24000, 32000, 44100
- **bitrate**: integer (default: 256000) - Bitrate for the generated music
  Options: 32000, 64000, 128000, 256000
- **audio_format**: string (default: mp3) - Output format. Use wav for production work
  Options: mp3, wav, pcm

## Examples

### Folk Country Song
Nostalgic folk country song from structured lyrics
```json
{
  "model": "minimax/music-2.5",
  "input": {
    "prompt": "Folk country, fingerpicked acoustic guitar, warm male vocal, nostalgic, intimate studio feel",
    "lyrics": "[Verse]\nOld screen door don't close right anymore\nCreaks like my knees on a cold morning\n\n[Chorus]\nBut time don't ask permission\nAnd rivers don't run backwards, friend\n\n[Outro]\n(Harmonica fades out)",
    "audio_format": "mp3",
    "bitrate": 256000,
    "sample_rate": 44100
  }
}
```

### Instrumental with [Inst] Tags
Instrumental track using [Inst] tags and parenthetical directions
```json
{
  "model": "minimax/music-2.5",
  "input": {
    "prompt": "Cinematic orchestral, 70 BPM, sweeping strings, gentle piano, emotional film score",
    "lyrics": "[Intro]\n(Soft piano)\n\n[Inst]\n(Strings swell)\n\n[Outro]\n(Piano fades)",
    "audio_format": "wav"
  }
}
```

## Response Format
```json
{
  "job_id": "abc123",
  "status": "pending | processing | completed | failed",
  "result": "URL or data (when completed)"
}
```

## Usage Flow
1. POST /predictions with model and input -> receive job_id
2. GET /predictions/{job_id} -> poll until status is completed or failed
3. Result contains output URL(s)

## Tips
- lyrics is required (1-3,500 chars) — write structured lyrics with [Verse]/[Chorus] tags for best results
- Prompt order: genre, mood, vocal description, BPM, instruments, era reference, production style
- Name specific instruments and vocal traits (gender, timbre, delivery) instead of generic terms
- For instrumentals, use [Inst] tags and parenthetical directions like '(Strings swell)' instead of words
- English and Mandarin vocals are strongest; keep wav + 44100/256000 for production
- For auto-generated lyrics or the dedicated instrumental mode, use Music 2.6 instead

## Documentation
https://replicate.com/minimax/music-2.5