Workflow setup and model selection

0:00 welcome to episode 39 of our comfy UI tutorial Series today I will show you how to use the WAN 2.1 model along with some lorus to unlock popular effects you see on social media such as squish inflate Crush rotate and others I covered the W model in more detail in episode 36 so I won't spend too much time on it in this one or go into too many details open comfy UI go to the workflow menu and choose open then you can load a workflow as usual all my workflows are free to download download from Discord let's start with a popular effect like the squish effect what's different compared to the previous episode is this Laura node which allows us to load Aura for the onean model as always I've included everything you need to get started in this node you can download the models you need by clicking on the download link and there's information on which folder to place them in uh the same applies for the clip mode vision and vae each workflow from this episode will have the same structure with the only difference being the Laura that gives us the effect the only custom node you'll need is the video helper

1:09 node you can go to the manager search for video helper install the node and then restart comfy UI if you still have some missing nodes it's because you haven't updated comfy UI go to the manager and click on update all some have reported that the manager failed to update and still showed missing one nodes that come with the latest version of comfy UI in that case go to your portable comfy UI find a folder called update and run the update comfy ui.

1:32 bat file then reopen comfy UI and it should be good to go now we're back in confy UI and there are no red nodes so everything is working fine compared to the previous episode The Laura is different and the model I'm using is different uh the Laura is trained on a a 480p image to video model so I used this version you have more versions available if you have a powerful computer but this one is faster so I used it the other version is supposed to be better but is double the size and as you probably know W models are very slow even on my PC so you need