ACE-Step Complete (Base Only)

A ComfyUI node in Audio/ACE-Step with 26 inputs and 3 outputs.

By kana112233·Created 6 months ago·Updated 6 months ago· 27

◄add_drumstrue►

◄add_basstrue►

◄add_guitarfalse►

◄add_keyboardfalse►

◄add_stringstrue►

◄add_synthsfalse►

◄add_percussionfalse►

◄add_brassfalse►

◄add_woodwindsfalse►

◄add_backing_vocalsfalse►

◄add_fxfalse►

◄add_vocalsfalse►

◄vocal_languageunknown►

◄lyrics►

◄caption►

◄seed-1►

◄inference_steps50►

◄guidance_scale7.00►

◄use_adgfalse►

◄cfg_interval_start0.00►

◄cfg_interval_end1.00►

◄audio_formatflac►

◄checkpoint_diracestep-v15-base►

◄lm_model_pathacestep-5Hz-lm-1.7B►

◄deviceauto►

CategoryAudio/ACE-Step

Inputs (26)

Name	Type	Default	Description
src_audio	AUDIO		Source audio to complete.
add_drumsopt	BOOLEAN	true	Add drums track
add_bassopt	BOOLEAN	true	Add bass track
add_guitaropt	BOOLEAN	false	Add guitar track
add_keyboardopt	BOOLEAN	false	Add keyboard/piano track
add_stringsopt	BOOLEAN	true	Add strings track
add_synthsopt	BOOLEAN	false	Add synthesizer track
add_percussionopt	BOOLEAN	false	Add percussion track
add_brassopt	BOOLEAN	false	Add brass track
add_woodwindsopt	BOOLEAN	false	Add woodwinds track
add_backing_vocalsopt	BOOLEAN	false	Add backing vocals track
add_fxopt	BOOLEAN	false	Add FX/sound effects track
add_vocalsopt	BOOLEAN	false	Add vocals track
vocal_languageopt	COMBO	unknown	Language for vocals.
lyricsopt	STRING		Lyrics text.
captionopt	STRING		Style description.
seedopt	INT	-1-1–18446744073709550000	Random seed. -1 for random.
inference_stepsopt	INT	5020–100	Diffusion steps. Higher = better quality.
guidance_scaleopt	FLOAT	7.001–15	CFG scale.
use_adgopt	BOOLEAN	false	Adaptive Dual Guidance.
cfg_interval_startopt	FLOAT	0.000–1	CFG start ratio.
cfg_interval_endopt	FLOAT	1.000–1	CFG end ratio.
audio_formatopt	COMBO	flac	Output audio format.
checkpoint_diropt	COMBO	acestep-v15-base	Model directory.
lm_model_pathopt	COMBO	acestep-5Hz-lm-1.7B	Language model.
deviceopt	COMBO	auto	Compute device.