This is my commonly used pure music generation workflow, which has the advantage of being fast and can generate 2 minutes of pure music within 10 seconds. However, its disadvantage is that the sound quality is poor and it cannot handle high-quality large-scale projects.


Its basic settings:
TextEncodeAceStepAudio: Fill in the first column with prompt words such as music style and instrument, which are the key to the entire music.
EmptyAceStepLatentAudio: Set the duration of the music.


The download page for the model:
https://huggingface.co/Comfy-Org/ACE-Step_ComfyUI_repackaged/tree/main/all_in_one
Model placement location: comfyui \ models \ checkpoints