Original editorial illustration for this topic. Not an official product screenshot or a depiction of the exact reported event.
Runway is useful when you want to animate a visual concept rather than manually build every frame. Its controls and model availability change, so always confirm the selected model before generating.
Screenshot is unavailable in this browser. Open the original.Step 1 — prepare the source frame
Use a clear image with an obvious subject, clean edges and room for the motion you want. Image-to-video relies on the source image for its visual starting point. If a label must stay readable, keep it large and test for distortions.
Step 2 — choose a supported video model
Open Runway’s creation workspace and select a model supporting image-to-video. Runway’s official Gen-4.5 documentation describes both text-to-video and image-to-video controls. Your account, plan and the interface may differ from screenshots.
Step 3 — prompt the motion
A smooth, slow dolly-in toward the glass bottle on the desk. Reflections shift naturally as the camera moves. The bottle remains stationary, with consistent shape and label. Soft daylight, single continuous cinematic shot.
The best video prompt explains what changes over time: subject movement, environment movement and camera movement.
Step 4 — generate, inspect, revise
- Choose the intended clip length and aspect ratio.
- Generate a short test before making multiple variations.
- Review details such as hands, lettering, physics and sudden jumps.
- Revise one motion parameter, then render again.
- Export only after checking licensing and use rights in the current plan.
Common mistakes
Too many simultaneous camera directions make the scene hard to control. A short, physical description often works better than a long list of vague adjectives. Consult the official Image to Video Prompting Guide for examples.
Primary sources
Official vendor documentation was used to verify the workflow. Some options depend on plan, region, product version or account access.