Sing your favorite songs alongside AI! Our music is generated through different machine learning models to give you the most authentic singing experience, hand-picked by our team. By selecting either the classic MIDI karaoke feature or the experimental jukebox singalong feature, you can experience the bleeding-edge technology of OpenAI’s Jukebox and the creativity of Magenta.js’ Music Transformer!
To start, simply hit the “Get Started” button at the top of the page; the website will help you with the rest!
We trained each of the models with audio files, either raw audio or MIDI versions, to generate short clips of each of the songs. After some priming and a lot of waiting, we created music that you can sing along to! Be sure to check out our project’s documentation to learn more about the process, as well as how things came to be!