Generative Video

Building repeatable text- and image-to-video pipelines around Wan models inside ComfyUI — testing motion coherence, prompt control and render cost for practical short clips.