riffusion/riffusion

Stable diffusion for real-time music generation

Capabilities

No capability data available

Community model (estimated from hardware time)

Name	Type	Description	Default	Constraints
`alpha`	number	Interpolation alpha if using two prompts. A value of 0 uses prompt_a fully, a value of 1 uses prompt_b fully	`0.5`	min: 0, max: 1
`denoising`	number	How much to transform input spectrogram	`0.75`	min: 0, max: 1
`num_inference_steps`	integer	Number of steps to run the diffusion model	`50`	min: 1
`prompt_a`	string	The prompt for your audio	`"funky synth solo"`	—
`prompt_b`	string	The second prompt to interpolate with the first, leave blank if no interpolation	`—`	—
`seed_image_id`	string	Seed spectrogram to use	`"vibes"`	agilemarimmask_beat_lines_80mask_gradient_darkmask_gradient_top_70mask_graident_top_fifth_75mask_top_third_75mask_top_third_95motorwayog_beatvibes

alphanumber

Interpolation alpha if using two prompts. A value of 0 uses prompt_a fully, a value of 1 uses prompt_b fully

Default: 0.5min: 0, max: 1

denoisingnumber

How much to transform input spectrogram

Default: 0.75min: 0, max: 1

num_inference_stepsinteger

Number of steps to run the diffusion model

Default: 50min: 1

prompt_astring

The prompt for your audio

Default: "funky synth solo"

prompt_bstring

The second prompt to interpolate with the first, leave blank if no interpolation

seed_image_idstring

Seed spectrogram to use

Default: "vibes"

agilemarimmask_beat_lines_80mask_gradient_darkmask_gradient_top_70mask_graident_top_fifth_75mask_top_third_75mask_top_third_95motorwayog_beatvibes

Version: 8cf61ea6c56aUpdated: 8/13/20261.1M runs