Can Suno Separate split a song I uploaded?

No. Suno Separate only works on tracks made by a Suno node, because it finds the song by its Suno task ID and audio ID. For any other song, use Audio Separation.

What are the Task ID and the Audio ID?

They identify a song made with Suno. When a Suno node is connected, both are taken from it automatically at run time, so you do not need to type them.

How many stems does the 12 Stems mode return?

Up to 12, such as drums, bass, guitar, piano and vocals. The stems you get depend on what the model finds in the song.

Should I choose Vocal / Instrumental or 12 Stems?

Start with Vocal / Instrumental when you only need the vocals or the backing track. Choose 12 Stems when you need each instrument on its own; it takes longer and costs more.

Suno Separate node — split a Suno song into vocals, instrumental or 12 stems

Split a song made with Suno into a vocal and an instrumental track, or into up to 12 stems such as drums, bass, guitar and piano, straight from a Suno node.

The Suno Separate node splits a song made with Suno into its parts. Vocal / Instrumental returns a vocal track and an instrumental track. 12 Stems splits the song into up to 12 stems, such as drums, bass, guitar and piano. The node works on tracks from Suno nodes only; for any other audio, use Audio Separation.

When to use it

Take the vocals out of a generated song to remix or sample them.
Keep the instrumental for karaoke or as background music.
Split a song into stems to remix or master it.
Prepare the vocals for a voice change or a dub.
Make an acapella version of a generated song.

Quick start

Add the node

Press Tab on the canvas and choose Audio › Clean & Separate › Suno Separate.

Wire a Suno song

Wire a Suno node, such as Suno Create Music, into the Audio input. The song's Task ID and Audio ID are taken from it at run time.

Choose the separation

Open the settings panel and set Separation Type to Vocal / Instrumental or 12 Stems.

Run it

Click Run on the node. The parts appear in the settings panel, each with its own player.

A generated song is split: the vocals get a new voice, and the instrumental gets new vocals.

Input and outputs

Input	Accepts	What it does
Audio	Suno nodes, such as Suno Create Music	The song to split. The node reads its Suno task ID and audio ID.

Output	What it carries
Vocals	The isolated vocal track.
Instrumental	The isolated instrumental track.

With 12 Stems, each stem appears in the settings panel with its own player. The stem names depend on what the model finds in the song.

Settings

Setting	What it does
Separation Type	Vocal / Instrumental (the default) returns two tracks. 12 Stems returns up to 12 stems.
Task ID	The Suno task ID of the song. Taken from the connected Suno node at run time.
Audio ID	The Suno audio ID of the song. Taken from the connected Suno node at run time.

A connected Suno node always wins over IDs typed by hand. Type the IDs only when no Suno node is connected, for example to split a song from an earlier workflow.

Credits

12 Stems costs more than Vocal / Instrumental, and it takes longer.

Tips

Start with Vocal / Instrumental. Move to 12 Stems only when you need each instrument separately.
Re-voice the singer. Wire the vocals into Voice Changer or Dubbing.
Write new vocals. Wire the instrumental into Suno Add Vocals to sing new vocals over the backing track.
Songs from elsewhere. For a song that was not made with Suno, use Audio Separation instead.

From the API

The MCP tool is suno_separate_stems. See the MCP tools.

Suno Separate