# Generate Music

> Create original music with MiniMax Music from a reference song, voice or instrumental plus a text prompt, with optional lyrics, genre and mood.

Source: https://nodaro.ai/docs/nodes/audio/generate-music

The **Generate Music** node creates an original music track with MiniMax Music from a reference song, voice or instrumental plus your text description. The reference sets the direction: its genre, instrumentation and energy. Your prompt describes what should be different. For music from a text prompt alone, use [Suno Create Music](https://nodaro.ai/docs/nodes/audio/suno-create-music).

- Found in: Audio › Music
- Output: audio
- API type: `generate-music`

**A reference is required:** 
MiniMax Music does not run without a reference song, voice or instrumental. When no reference is set, Nodaro refuses the run before any credits are reserved, in the editor, in workflow runs and over the API.

## When to use it
- You have a track whose feel you like, and you want a new, original piece in the same direction.
- You need background music for a video, a jingle or a short musical sting.
- You need an instrumental bed for a podcast or a voiceover.
- You want to try music ideas before you work with a composer.

## Quick start
### Add the node

Press Tab on the canvas, or open the node picker in the sidebar, and choose **Audio › Music › Generate Music**.

### Choose the model

Open the settings panel and choose **MiniMax Music** under **Provider**. The reference and lyrics settings appear once MiniMax Music is selected.

### Add a reference

Connect an audio node, such as [Reference Audio](https://nodaro.ai/docs/nodes/audio/reference-audio) or [Upload Audio](https://nodaro.ai/docs/nodes/audio/upload-audio), to the **Ref audio** input. You can also set the reference in the settings panel, under **Reference Audio**.

### Describe the music

Write the **Prompt**, for example `Upbeat electronic dance track, 120 BPM, synth leads, punchy drums, building energy`.

### Run it

Click **Run**. The track appears on the node, and every earlier result stays in its results.

Workflow: A reference track, a prompt and a Music Mood picker feed Generate Music. The new track becomes the soundtrack of a video.

- Reference Audio → Generate Music (ref audio)
- Text → Generate Music (prompt)
- Music Mood → Generate Music (audio style)
- Generate Music → Merge Video & Audio (audio)

## Inputs
| Input | Accepts | What it does |
| --- | --- | --- |
| **Prompt** | Text nodes and pickers | The description of the music. |
| **Ref audio** | Audio nodes, such as Reference Audio, Upload Audio and Suno Create Music | The reference track. A connected reference is used instead of the one in the settings panel. |
| **Audio style** | [Music Genre](https://nodaro.ai/docs/nodes/creative-controls/music-genre), [Music Mood](https://nodaro.ai/docs/nodes/creative-controls/music-mood) and [Instrumentation](https://nodaro.ai/docs/nodes/creative-controls/instrumentation) pickers | Music Genre supplies the genre and Music Mood supplies the mood. Instrumentation adds its choice to the prompt, and an instrumental choice turns off the vocals. |

The output, **Audio**, is the generated track. When pickers are connected, the settings panel shows the **Final prompt** that will be sent.

## Settings
| Setting | What it does |
| --- | --- |
| **Provider** | The music model. **MiniMax Music** is the only choice. |
| **Prompt** | The description of the music: genre, mood, tempo, instruments and structure. You can insert the value of another node with a variable, such as `{Theme}`. See [Prompt variables](https://nodaro.ai/docs/concepts/prompt-variables). |
| **Lyrics** | Optional words for a sung track. Line breaks mark the phrasing. |
| **Reference Audio** | Where the reference comes from: **None**, **Upload file** or **YouTube URL**. An uploaded file can be MP3, WAV, M4A, AAC or FLAC. For a YouTube link, paste the address and click **Get**. When an audio node is connected to **Ref audio**, the panel shows that node instead. |
| **Genre** | An optional genre, such as `rock`, `jazz` or `electronic`. |
| **Mood** | An optional mood, such as `upbeat`, `melancholic` or `epic`. |
| **Instrumental (no vocals)** | Makes a track without vocals. It is on by default. |
| **Pre & post text** | Text that is always added before and after the prompt. It is hidden from people who use your workflow as an app. See [Prompt pre and post text](https://nodaro.ai/docs/concepts/prompt-pre-post-text). |

## How the reference shapes the result

- **Choose a close reference.** MiniMax Music builds on the reference. Choose one whose genre, instrumentation and energy are close to what you want.
- **Describe the difference.** Use the prompt for what should change: another tempo, other instruments, a different mood.
- **The result is never a copy.** The reference influences the mood and the instrumentation, not the melody.

## Tips
- **Write a detailed prompt.** `Upbeat electronic dance track, 120 BPM, synth leads, punchy drums, building energy` works better than `dance music`.
- **Keep Instrumental on for backgrounds.** Instrumental tracks are easier to place under video and voiceover.
- **Put the track under a video.** Connect the result to [Merge Video & Audio](https://nodaro.ai/docs/nodes/video/merge-video-audio), or layer it with other sounds in [Mix Audio](https://nodaro.ai/docs/nodes/audio/mix-audio).
- **Use the Suno nodes for song work.** For covers, extensions, stems and section edits, start from [Suno Create Music](https://nodaro.ai/docs/nodes/audio/suno-create-music).

## Frequently asked questions

### Why does Generate Music need a reference track?

The node runs MiniMax Music, which builds its music on a reference song, voice or instrumental. A run without a reference is refused before any credits are reserved. For music from a text prompt alone, use Suno Create Music.

### How many credits does Generate Music cost?

Each run costs 180 credits. A run that is refused because it has no reference costs nothing.

### Does the result copy the reference song?

No. The reference shapes the mood and the instrumentation, not the melody, so the result is an original track.

### Can Generate Music make a song with vocals?

Yes. Write lyrics in the Lyrics field and turn off Instrumental (no vocals), which is on by default.

### What is the difference between Generate Music and Suno Create Music?

Generate Music runs MiniMax Music and needs a reference track. Suno Create Music runs Suno, needs only a prompt, and works with the other Suno nodes that extend, cover and edit songs.