Overview
MiniMax Music 2.0 by Minimax AI
Designed for developers and creative teams, it produces complete songs that feel composed and arranged rather than mechanically generated.
MiniMax Music 2.0 API Overview
MiniMax Music 2.0 is a generative audio model that converts descriptive prompts and lyrics into full musical compositions. Instead of focusing on short clips or isolated loops, it delivers cohesive tracks with a clear beginning, progression, and resolution.
The system interprets both creative intent and structural cues, meaning that when a user provides lyrics formatted with sections like verses and choruses, the model reflects that structure directly in the output. This creates a more predictable and controllable generation process, especially valuable in production environments.
Technical Architecture
The architecture is designed to align linguistic meaning with musical expression. By processing text and audio relationships simultaneously, the model ensures that lyrics match rhythm, melody, and phrasing in a natural way.
| Component | Description |
|---|---|
| Architecture | Transformer-based model optimized for audio generation |
| Input Types | Prompt (style control) + lyrics (structure and vocals) |
| Output | Fully composed audio track with vocals and instruments |
| Generation Method | Joint modeling of language and sound |
Output Specifications
| Feature | Value |
|---|---|
| Maximum duration | Up to ~5 minutes |
| Audio quality | 44.1 kHz stereo |
| Format | Compressed audio (e.g., MP3) |
| Bitrate | Up to 256 kbps |
| Structure | Full song with defined sections |
Core Capabilities and Creative Control
Structured Music Generation
MiniMax Music 2.0 operates through a dual-conditioning mechanism. A descriptive prompt shapes the overall sound, defining genre, mood, tempo, and instrumentation, while lyrics guide the vocal line and narrative structure. This separation allows for precise creative direction without requiring technical audio expertise.
Unlike many AI music systems that treat input as a loose suggestion, this model follows structural intent closely. Sections such as intros, verses, and choruses are not only recognized but translated into meaningful musical transitions, preserving flow and continuity throughout the track.
Expressive Audio Output
One of the defining characteristics of the model is its ability to generate expressive vocals paired with well-balanced instrumentation. The vocal delivery carries tonal variation and emotional nuance, while the instrumental layer adapts dynamically to support the progression of the song.
This balance results in outputs that resemble produced tracks rather than synthetic experiments. Genres can shift naturally depending on the prompt, allowing the same system to generate anything from soft acoustic arrangements to high-energy electronic compositions.
Long-Form Composition
MiniMax Music 2.0 supports extended generation, producing tracks that can reach up to approximately five minutes in duration. This capability enables complete storytelling within a single output, making the model suitable for real-world media usage where continuity matters.
Generation Code Sample
Output Code Sample
Real-World Applications
Content and Media Production
MiniMax Music 2.0 fits naturally into modern content pipelines, where speed and consistency are critical. It can generate background music for videos, podcasts, and advertising campaigns while maintaining a cohesive style across multiple outputs. This makes it especially valuable for teams producing high volumes of media who need reliable, on-demand audio without compromising quality.
Developer Integrations and Platforms
For developers, the API enables seamless integration into a wide range of creative platforms, including music applications and AI-powered editing tools. Its structured input approach ensures predictable and repeatable results, which is essential when building user-facing features that rely on consistency and control.
Rapid Prototyping for Musicians
At the same time, musicians and producers can use the model as a rapid prototyping tool. It provides a fast way to explore musical ideas, experiment with different genres, and generate vocal drafts without the need for recording sessions. This significantly accelerates the early stages of the creative process while reducing production overhead.
Comparison with Other Models
vs Suno Music: MiniMax Music 2.0 excels in longer track generation up to 5 minutes with detailed instrument separation, while Suno produces shorter tracks faster and focuses on radio-ready pop style with highly accessible vocal synthesis.
vs Stable Audio 2.0: Stable Audio uses diffusion-based methods focusing on experimental sound design and precise sonic control. MiniMax Music 2.0 contrasts with more conventional song structures and emotional vocals, making it more suited for commercial music production.
vs Soundverse: Soundverse is known for its comprehensive toolset including stem separation and auto-complete features, catering to both hobbyists and professionals. MiniMax matches Soundverse in audio quality, but stands out with its patented vocal synthesis and longer track generation up to 5 minutes.
Pricing
| Usage | Price | Unit |
|---|---|---|
| Usage | $0.03 | per generation |
Quick start
View docs (opens in a new tab)
Build with your agent
Give your agent the model details, code and documentation.
1Get your API key
Create API key (opens in a new tab)export AIMLAPI_KEY="8f2b4c7d9a1e6b3f…"2Make your first request
curl 'https://api.aimlapi.com/v2/generate/audio' \
-H "Authorization: Bearer $AIMLAPI_KEY" \
-H 'Content-Type: application/json' \
-d '{
"model": "minimax/music-2.0",
"prompt": "An upbeat lo-fi hip hop instrumental with warm piano chords, soft vinyl crackle and a relaxed drum groove",
"lyrics": "[Verse]\nStreetlights flicker, the night breeze sighs\nShadows stretch as I walk alone\nAn old coat wraps my silent sorrow\nWandering, longing, where should I go\n[Chorus]\nPushing the wooden door, the aroma spreads\nIn a familiar corner, a stranger gazes"
}'import os
import requests
headers = {
"Authorization": f"Bearer {os.environ['AIMLAPI_KEY']}",
"Content-Type": "application/json",
}
payload = {
"model": "minimax/music-2.0",
"prompt": "An upbeat lo-fi hip hop instrumental with warm piano chords, soft vinyl crackle and a relaxed drum groove",
"lyrics": "[Verse]\nStreetlights flicker, the night breeze sighs\nShadows stretch as I walk alone\nAn old coat wraps my silent sorrow\nWandering, longing, where should I go\n[Chorus]\nPushing the wooden door, the aroma spreads\nIn a familiar corner, a stranger gazes"
}
response = requests.post("https://api.aimlapi.com/v2/generate/audio", headers=headers, json=payload, timeout=120)
response.raise_for_status()
print(response.json())const response = await fetch("https://api.aimlapi.com/v2/generate/audio", {
method: "POST",
headers: {
Authorization: `Bearer ${process.env.AIMLAPI_KEY}`,
"Content-Type": "application/json",
},
body: JSON.stringify({
"model": "minimax/music-2.0",
"prompt": "An upbeat lo-fi hip hop instrumental with warm piano chords, soft vinyl crackle and a relaxed drum groove",
"lyrics": "[Verse]\nStreetlights flicker, the night breeze sighs\nShadows stretch as I walk alone\nAn old coat wraps my silent sorrow\nWandering, longing, where should I go\n[Chorus]\nPushing the wooden door, the aroma spreads\nIn a familiar corner, a stranger gazes"
}),
});
if (!response.ok) throw new Error(`HTTP ${response.status}`);
console.log(await response.json());Related models
Frequently asked questions
What inputs and outputs does MiniMax Music 2.0 support?
MiniMax Music 2.0 takes audio, text as input and returns audio.
When was MiniMax Music 2.0 released?
MiniMax Music 2.0 became available on December 5, 2025.
How is MiniMax Music 2.0 billed?
MiniMax Music 2.0 is billed per generation, a fixed charge per output rather than by prompt length.
Who created MiniMax Music 2.0?
MiniMax Music 2.0 was built by MiniMax.
How do I call MiniMax Music 2.0 on AI/ML API?
Send a request to v2/generate/audio with minimax/music-2.0 as the model id.
Is MiniMax Music 2.0 available through a single API key?
Yes. MiniMax Music 2.0 is served through AI/ML API, so the same key and endpoint format used for other models applies.