Soniqo MusicFree app

Make karaoke tracks. For free.

Make karaoke backing tracks from your own audio. Remove vocals, split stems, change key and export WAVs for free on Android, iPhone and iPad.

All current tools and WAV exports are free. No account, ads or subscription.

Get it on Google Play Download on the App Store

Read the app guide Explore the tools

Available on Google Play and the App Store.

Android 10+ (64-bit) · iPhone and iPad with iOS 18+

How to make a karaoke track from a song

Import your audio
Choose your own WAV, M4A or MP3 file, or try the bundled sample.
Remove the vocals
Open Split vocals & instruments, choose a section or the whole recording, then select Instrumental.
Keep your backing track
Listen to the result and export a WAV. You can also change the key or speed of a section up to 10 minutes.

Follow the setup and export guide

Explore the tools

Remove vocals

Choose "Instrumental" for the music without the vocal part. Listen or export a WAV. Choose Vocals, Drums, Bass or Other for an isolated part.

Isolate instruments

Separate vocals, drums, bass and other instruments. Listen to an individual part or keep it as a WAV for another project.

Change speed and pitch

Speed ranges from 0.25× to 2×, and pitch from −12 to +12 semitones. A/B markers select a section up to 10 minutes for these edits. Export saves the prepared WAV, up to 1 hour including repeats; use Save to Files or another destination in the share sheet.

Clean voice recordings

Reduce background noise and room reverb in speech after downloading a separate cleanup model once (about 301 MB on Android, or 427 MB on iPhone and iPad). Compare the original and cleaned audio before saving a WAV. Cleanup of singing is experimental and can change the voice.

See the app

Screenshots show the English iPhone app. Separation can leave some vocal or instrument leakage.

Karaoke maker questions

Is the karaoke maker free?: All current tools and WAV exports are free, with no account, ads or subscription.
Does it show lyrics or record my singing?: The app creates and edits backing-track audio from your own files. It does not include synced lyrics, a song catalogue or singing recording, and it does not download songs from streaming services.
Can I remove vocals offline?: Yes, after a one-time separation model download: about 305 MB on Android or 233 MB on iPhone and iPad. Keep the app open while processing. Separation can leave some vocal or instrument leakage.

Duration and storage

The first split downloads a model: about 305 MB on Android, or 233 MB on iPhone and iPad. Allow internet access and free storage for setup. Later splits work offline with the cached model.

Separate recordings up to 2 hours, or choose a shorter section from a longer file. Keep the app open while processing; going to the background cancels the operation. Long recordings need substantial free storage. The app shows a storage estimate before starting.

Audio and model files

Audio processing runs on your device. Model setup downloads data from Hugging Face; it does not upload your recording. Remove downloaded models in the app. Save a WAV to your files for a separate copy outside the app. All current tools are free, with no account or subscription.

Read the app guide