To be used with https://github.com/Comfy-Org/ComfyUI/pull/16825.

This repository's VAE file contains only the decoder part. This is lightweight and fine for most cases.

However, if you would like to use your own soundtrack (e.g. like in this workflow), to encode the audio, you will need to download the full version at https://huggingface.co/kabachuha/kandinsky6_audio_vae_full.

Downloads last month
115
Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐Ÿ™‹ Ask for provider support

Collection including kabachuha/kandinsky6_audio_vae