I am using exclusively comfyui, but i think the other platforms will or have incorporated the model soon. The bigger Problem is the VRAM demand of the model. I am not able to run the DEV Model on my 3080 locally. At least not efficient. I´ve switched to a Cloud provider, runpod, and i am renting an A40 48 GB VRAM to generate the base images. Prices are relatively cheap with .35 $ per hour. Also you can rent persistent storage, so that you have not to set up your comfy enviroment new after you killed your current pod.