This ComfyUI workflow turns a single still image into a short, dynamic video using Runway’s Gen4 Turbo image-to-video model. It wires together four nodes: LoadImage to bring your source image into the graph, RunwayImageToVideoNodeGen4 to call the hosted Gen4 Turbo API and synthesize motion over time, SaveVideo to encode the result to a standard video file, and a MarkdownNote with links and setup guidance. Because the heavy lifting runs through the Runway partner node, the workflow stays lightweight and fast on your local machine.

Technically, the LoadImage node supplies the pixel data to RunwayImageToVideoNodeGen4, which sends the image (and your chosen settings like duration and frame rate) to the Runway API. The API returns a generated video stream compatible with ComfyUI’s SaveVideo node, which encodes the final clip (e.g., MP4) at your selected FPS. The partner node uses HTTPS and requires a secure network setup as noted in the Author Notes (see "About API Node"): run from localhost/127.0.0.1 and ensure outbound HTTPS is allowed. The result is a clean, reproducible image-to-video pipeline that’s easy to iterate: swap the input image, adjust motion/duration parameters, and re-run to explore different looks.