We have created a unique online application for separating songs into vocals and music using neural networks. Our MDX and UVR models are trained using our own dataset, which includes music of various genres.
High quality is ensured by resource-intensive algorithms, which we use powerful and expensive servers. Your financial support helps not only to pay for hardware rental, but also to develop the project further.
We offer both free and premium versions to our users. In the second case you get:
1. Bigger limits (depending on the subscription level). The more audio minutes, the more songs you can process.
2. WAV / FLAC downloading.
3. Backing vocals extractor.
4. Piano and strings models.
As an experiment, these premium benefits will be available to all users for a while:
1. Better quality processing with the latest models.
2. Separation into lead vocal, bass, drums, guitar or saxophone.
If you are a new subscriber, you will receive a link in a private message immediately after payment. Or you can find it here: