Why is vocal separation slow in my browser?
The short answer
Browser separation has several stages: downloading a model, decoding your audio, running the model and preparing downloads. A slow first run may be mostly download time. Once processing begins, song length, hardware and memory pressure determine how quickly it can finish.
Try this workflow
- Start with Fast AI and a ten-second preview.
- Keep the tab open and close other demanding apps.
- Use a shorter clip if the device cannot comfortably process the full song.
What to check before you finish
Check the status message to distinguish downloading from separating. Avoid repeatedly clicking Process or opening several jobs at once. Cancel an unwanted job before trying another method. There is no universal completion time that applies to every phone or computer.
A practical example
If the status says models are downloading, let that stage finish once rather than starting another job. If it says separating and your computer becomes unresponsive, cancel and try a short clip. This tells you whether the device can handle the workload without committing to another long wait or creating competing jobs.
More questions
Does a faster internet connection speed up separation?
It helps model downloads. Once the model is running locally, processor speed and memory matter more than internet bandwidth.
Does closing the tab let processing finish elsewhere?
No. This workflow runs on your device, with no background AI server finishing the job for you. Keep the tab open until the result is ready.
Free tools · No account required. Download the results you want to keep.