After hooking up local AI video into ComfyUI, I’ve been thinking about where game dev will use it first. My take is concept phase and cutscenes will get the biggest boost first. Early in game development you need tons of atmosphere demos, character motion references, and scene camera drafts—stuff that used to rely on outsourcing or manual keyframing.
Now with local video nodes you can crank out a few versions, let the director quickly preview and lock in direction, basically zero cost. Formal assets won’t make it in short-term—resolution, temporal consistency, and controllability aren’t ready for the actual pipeline yet. But as a pre-vis and communication tool, it already saves a ton of back-and-forth.
The beauty of ComfyUI is you can string together image generation, video generation, and post-processing into your own custom chain, and the team can tweak it as needed. Using it for “quick idea validation” is way more realistic than forcing it into the final pipeline.