YES

As of 13 August 2026, AI can create a karaoke version of a song.

Most people should do this themselves; the prompt is on this page.

Can you do it?

15 minutesto a draft.

30 minutesto something you’d act on.

Cost, all in£0 one-off

Skill neededchat-fluent

Who has to check ityou

What the alternative costsA manual audio editor is the alternative; no price is stated here.

If this goes wrong: some vocals, reverb or backing instruments remain in the track, so you replace the source or adjust the separation and try again.

What to actually do

  1. Do it yourself

    The route this page recommends

    A chat interface, chat-fluent skill, and roughly 30 minutes until you can act on the result.

    How to actually do it

    1. Open LALAL.AI and check that you have permission to process the song and use the resulting file for your intended purpose.
    2. Gather the highest-quality audio file you have, preferably the original download rather than a recording of a speaker, and upload it to the stem-splitting tool.
    3. Select vocal and instrumental separation, then choose the instrumental stem or vocal-removal option rather than a full song-generation feature.
    4. Preview the instrumental result from the beginning, chorus, quiet sections and any parts with backing vocals, then adjust the separation settings if the music sounds damaged or vocals remain.
    5. Export the instrumental file, play it on the device you will use for karaoke, and compare its volume, timing and sound against the original before using or sharing it.

    Prompt

    I want to create a karaoke version of the attached song, using only audio I have permission to process. Remove the lead vocals while keeping the instrumental accompaniment as natural as possible. If complete removal would damage the music, prioritise a clean-sounding instrumental over aggressive vocal removal. Export the result as a commonly playable audio file, and tell me what artefacts or remaining vocals I should listen for when checking it. Do not add, rewrite or recreate any part of the song.

    Open it prefilled in ChatGPT or Claude, or copy it into Gemini, which takes no prefill link.

  2. Use a tool built for this

    Second choice
  3. Hand it to a person

    The distant third

    A person who owns the outcome does this end to end, worth it when the failure is dear.

What it gets wrong

Even on a YES, the friction has a name: context depth and stakes of error.

How we scored this

Five axes, each scored nought to two by hand: ten means AI carries the task cleanly, and the thresholds that turn a total into YES, PARTLY or NO are published in the methodology. Each axis name links to its definition.

AxisScore (0–2)
Output2
Inputs2
Verification2
Liability1
Effort delta2
Total9 / 10

FAQ

Can AI remove vocals from a song?
Yes. AI stem-splitting tools can extract the vocal and instrumental parts of an uploaded track, but the result may contain vocal bleed or damage to the music.
What is the best AI tool for making karaoke tracks?
LALAL.AI is a suitable purpose-built option because it separates vocals and instruments from an audio track. Listen to the result before relying on it, as separation quality varies between recordings.
Can AI remove vocals without ruining the music?
Sometimes, but not consistently. Vocal and instrumental sounds can overlap, so a cleaner result may leave some vocals behind while more aggressive removal can make the accompaniment sound artificial.
Is it legal to make a karaoke version of a song with AI?
That depends on your permission and what you do with the resulting file. Making a private copy and sharing or performing it can involve different copyright issues, so seek advice from a UK copyright solicitor for a serious or commercial use.

Nearby answers

Assessed by gpt-5.6-luna (gpt-5.6-luna) on 2026-08-13, second-checked by an independent model. Wrong somewhere? Email [email protected] and it gets re-checked.

The newsletter

AI news, new answers and product picks, straight to your inbox.