How to Split Audio Files With – an AI-Powered Audio Splitter

Written by: Cody

Updated: Jun 17, 2022

If you’re not a producer, DJ, or someone who loves tinkering with audio files then chances are you haven’t heard of an audio splitter. On the other hand, if you are someone who knows what an audio splitter is then you’re likely aware of how useful they can be to split apart vocals from instrumentation.

I’ve recently become aware of a fairly new audio splitter called which has a pretty unique component to it. So today, I’ll be diving into how works, how to use it, and more.

What Is is an AI-driven audio splitting service that has the ability to separate vocals from the instrumentation of a song. It uses the power of artificial intelligence to learn how to better separate vocals from instruments over time. The more it’s used, the better it gets.

It’s really quite fascinating how complex and sophisticated a service has to be in order to accurately isolate and separate vocals from a song that has already been mixed and mastered.

According to’s blog:

Splitting just a single second of an audio track takes hundreds of millions of mathematical operations. Considering that an average song is about three minutes long, separating its vocal and instrumental tracks induces billions of math operations.

Yes, you read that right, billions of mathematical operations. And to consider that you can upload an average-sized audio file to and it will return the separated tracks in a matter of minutes is pretty impressive.

There’s TONS of data involved when it comes to training a neural network like’s. They state on their website that they’ve already processed 20TB worth of training data in order to train their system on how to perform the separation of instrumentation and vocals properly.

It’s safe to say, it has processed a few songs.

How to Use

Although their system is complex on the back end, is actually very simple to use. All you really need to do is choose a processing level and drag and drop your audio file onto their website. That’s it.

Then, just wait a minute or two and both a separated vocal track and instrument track will be rendered. offers 3 different processing levels:

  • Mild – Minimum processing, you may still hear some cross-over between tracks
  • Normal – Moderate processing, this is the default and should suffice in the majority of cases.
  • Aggressive – Maximum processing, heightened sensitivity and may lead to more artifacts however can be used if mild and normal aren’t enough

There isn’t a one-size-fits-all solution to choosing a processing level. offers 3 options due to the fact that every song is different in terms of mix and therefore using one processing level over another could return better results.

To demonstrate what can do. Have a listen to the following song by Lawrence Trailer entitled “Lets Call It”.

Now, we’ll run it through’s music separation service and we’ll get one track of vocals and one track of instrumentation. The song files below are the result:

Just the Vocals

Just the Instruments

Pretty impressive. Although you can still hear some artifacts (primarily in the instrumentation), this service will only continue learning and improving.

Additionally, in most music production and DJ scenarios, you typically won’t be using the whole song to make a remix. So, if some parts of the song are better split than others then that may not be an issue.

If it is an issue, try using a different processing level and/or uploading a higher quality version of the song to see how the splitting results differ.

Why Use a Song Splitter?

There are a few reasons why you might want to use a song splitter.

DJ/Music Producer

If you’re a DJ or music producer, using the best DJ software and song splitters are essential. They allow you to sample certain instrumental or vocal parts of a song so that they can be overlayed onto something else.

Without a song splitter, you would either have to find an original copy of an isolated track or attempt to re-create it yourself – both not easy to do.

A song splitter like can separate the vocals from the instruments in a matter of minutes leaving you tons of time to start working on a remix. Just imagine the possibilities that could come from having the ability to work with isolated tracks from any song you can think of.

Learning Purposes/Karaoke

Although most people will likely use a song splitter for commercial purposes, I think they’re also a great tool for learning purposes. Just think, whether you’re learning how to sing or you’re an instrumentalist learning how to arrange the instrumentation you can re-create a song with accompaniment.

If you’re a vocalist who wants to practice singing over their favorite song you can use the instrumental track and sing over it. Heck, it could also be used for karaoke.

On the other hand, if you don’t sing but play instruments you can use the vocal track and build your own instrumental section. With this technology, you can completely change how a song sounds, which just opens a whole new world of creativity.


There are 3 pricing options for

  1. Their first “Lite” option allows you to separate 10 audio tracks or 90 minutes worth of audio (whichever comes first) and is $10. This option is great if you are a hobbyist and want to play around with a few tracks. It’s inexpensive and allows you to download the full version of each separated track. You can still preview a song to see how well separates it before downloading the full version.
  2. Their second package is called the “Professional” and allows you to separate 30 tracks or 500 minutes of total audio. This is ideal for professional content creators such as DJ’s, music producers, etc.
  3. Their third and final package option is for high volume customers who need to process a large number of audio files. For this, contact directly to discuss a custom quote.

Overall, I think’s pricing is fair. The fact that they allow you to preview a track’s separated files before downloading the full tracks is ideal. Moreover, having the ability to use various processing levels and being able to separate a track’s vocals from its instrumentation in a matter of minutes is worth the cost, in my opinion.


Overall, I really like the concept of I haven’t used any service like it in the past, however, I can definitely see why it’s needed to make certain jobs easier.

We’re living in exciting times and to think that even a song-splitting service is continually getting smarter through the use of AI is something I think is pretty cool.

So if you have any need for a song separator, I’d recommend giving a test run and seeing how it goes. Play around with the different processing levels and determine if it’s a tool that will make your life/job easier.

If you’ve tried out let me know your thoughts in the comments section, I’m interested to see what others think.

About Cody
Cody is the founder of Musician Tuts, a free tutorial hub for musicians. He has over 15 years of experience playing a variety of instruments and dabbling in audio engineering. He spends his days blogging, listening to Spotify, and playing music.

You May Also Like…


Submit a Comment

Your email address will not be published.