Generate music from a text caption, lyrics, and style parameters.
API key obtained from the ModelBeam dashboard
Music description / style prompt
1 - 2000"AceStep_1_5_Turbo"
Song lyrics. Use "[Instrumental]" for no vocals
Duration in seconds
10 <= x <= 600Inference steps (8 for turbo, 32+ for base)
1 <= x <= 100Guidance scale
0 <= x <= 20Random seed (-1 for random)
Output format
"flac"
Beats per minute
30 <= x <= 300Musical key and scale, e.g. "C major"
Time signature
2, 3, 4, 6 Vocal language code
Reference audio for style transfer (max 10MB)
2048Music generation job created