Your songs will appear here
Describe a sound in the panel — each run returns about two full tracks, side by side on this canvas.
The examples all name genre, instruments and tempo. A music prompt that reads like a picture caption gets you a generic backing track.
One generation returns roughly two complete songs, so the choice is between finished tracks rather than another roll of the dice.
Lyrics and vocals come with the track — the model writes the words from your description and performs them.
Three steps, no setup
Genre, instruments, tempo, mood — or click an example.
About a minute later two tracks land on the canvas.
Play both, download the one that works.
Describe the sound rather than a scene — genre, instruments, tempo and mood. "Dreamy synthwave with warm analog pads, slow 90 BPM" gets you much closer than "a happy song".

Need moving pictures to go with it? Describe the scene and its motion.

Generate cover art for your track from a written description.


Describe the sound — genre, instruments, tempo, mood. The model writes the lyrics and sings them.
Try it free