Also we created a free guide so you can follow along:

Here is the full tutorial! You can find the free workflows and scripts in the video description & on my Patreon (no sign up required)

Add controllable AI creatures to your videos: FREE & OPEN SOURCE!. One input image is enough to create consistent characters and integrate them into your footage. Use Blender, After Effects, or open-source tools to control their movement. Full tutorial & free workflows below!

Here is the full tutorial! You can find the free workflows and scripts in the video description & on my Patreon (no sign up required)

For our newest AI short film we used 100% open source models to create consistent and controllable creatures!. We also used advanced video inpainting for two key shots. 👇Watch the short film below and see how this will change VFX! (Full tutorial and FREE workflows below)

"Create a side view of the pixar character with the slim head." (OmniGen2). Not a perfect result, but pretty promising. Still need to tweak a few settings. Hans Schmankerl is back!

"Change the brown jacket to blue." (OmniGen2)

Quick follow-up to yesterday’s workflow now in 3D! . Select bones in Blender, export as points, and drive the video with them using the ATI wan video model. More flexible than OpenPose since it works with any rig.

Here is the tutorial, you can find the link to the workflows in the video description (no sign up required)

Hunyuan 3D is amazing, but I always felt the textures could use a boost. So I built this workflow. Tutorial + free workflows below👇

RT @CitizenPlain: AE + Comfy process (and links to resources) here:

This is a fun workflow!. I animate a character using the Puppet Warp in After Effects, export the point data with @CitizenPlain 's method, and feed it into Kijai's Wan Video Wrapper using the new Wan Any Trajectory model.

Here is the tutorial with installation guides and free workflows. Download them from the link in the video description. No sign up required!

Had some fun improving 3D AI asset generation with open source models. Now supports importing existing 3D assets with UVs to generate 4K+ textures, or creating entirely new models from prompts/input images. Free Workflows + Tutorial below 👇

One of my favorite example shots from my last video! . Wan2.1 Start Frame + ControlNet

By the way, we have a website now! We'll put all the guides and additional resources on there in the future. It's an early version but feel free to take a look!

Full Tutorial & workflows here:

Our newest AI short film was 100% filmed on location. Then using open source video & image models in ComfyUI, we transformed me into a man who wakes up as a monstrous vermin!. Complete tutorial + FREE workflows shared below!

Updating the AI rendering workflow using Wan2.1. Pretty cool that you get extra movement like the wrinkles and the swinging rope!

Our newest AI Short Film “PAPER JAM”! . Each keyframe was animated by hand based on AI generated 3D models. "Rendered" using Flux + regional Loras in ComfyUI and interpolated using @Kling_ai