Help

Current steps and model choices

Which model should I choose?

Online models cover vocals/instrumental, four or six stems, instrument extraction, lead/backing vocals and dereverb. Check availability in the selector. Outputs are estimates, not named-speaker identification, general noise repair or guaranteed recovery of every sound.

After the full result is ready, compare the difficult passage with the original. Check both the wanted detail and unwanted leakage before choosing a track.

What is included?

Each track has independent Play/Pause, mute and volume controls. Shared playback, the common timeline and loop help compare the same passage. Playback adjustments do not create a new downloadable mix.

Compare completed results at the same timestamp and similar listening levels. Record the model, mode and detail to preserve. Trying another model requires a new task and uses another allowance or credits; it does not generate a free test excerpt.

Account

New Plus subscriptions are no longer sold. Existing confirmed subscriptions keep their paid-period rights; cancellation is managed in Account. These rights are separate from permanent credits.

Workflow wording reviewed on 6 November 2026. Models and account rules may change; use the live limits and availability shown here. Audio quality varies by recording.