Question 1

What stems can AI Stem Splitter separate?

Accepted Answer

It supports vocal/instrumental two-stem and multi-stem modes (vocals, drums, bass, guitar, piano, other). Available stems depend on the selected scene and model.

Question 2

What formats and sizes are supported?

Accepted Answer

We support MP3, WAV, FLAC, M4A, AAC, MP4, MOV, and other common formats. File size and duration limits vary by plan; long audio is best uploaded in segments.

Question 3

How is billing calculated?

Accepted Answer

Billable minutes = audio/video minutes × number of selected models. The system uses free quota first, then paid balance.

Question 4

How can I improve separation quality?

Accepted Answer

Prefer lossless or high-bitrate audio and avoid clipped distortion. Choose the right scene and model for the material; comparing multiple models can yield cleaner stem separation.

Question 5

How long are results kept?

Accepted Answer

Results are kept for a short period by default (typically 24–72 hours, and you can delete them immediately). After expiry, storage and indexes are cleared.

Question 6

Are my files secure?

Accepted Answer

Uploads use short-lived presigned links with encrypted transfer and access. The system keeps only necessary task metadata for history and troubleshooting, and you can delete it anytime.

Frequently Asked Questions

Still have questions?