Upload a start frame and end frame, describe the action and camera movement between them, choose your video settings, and continue generation in VisionStory.
Use a start and end frame to define where the shot begins, how it moves, and where it finishes.
Start, end, and motion
Compare the opening frame, closing frame, and generated motion side by side. The sequence makes it clear what stays stable and what changes as one action connects the shot.
The traveler's lowered hands, waiting dog, coastal sunset, and close framing establish the opening beat.
The raised cup defines the closing action while the traveler, dog, sunset palette, and camera angle remain recognizable.
The preview connects the lowered cup to the drinking pose as one readable action without changing the scene or subject identity.
Creative cue: keep identity, lighting, and framing stable; describe only the movement that must happen between the two anchors.
Frame pairing guide
A strong start-and-end-frame pair shares enough visual DNA to feel like one shot. Check the subject, camera logic, environment, and size of the change before generation so the model can spend its effort on motion instead of rebuilding the scene.
First and last frame video generation works best when the opening and closing compositions matter more than a completely open-ended prompt.
Plan the payoff before motion begins. A first frame can establish sealed packaging or a wider setup, while the last frame locks the hero angle, label visibility, and final arrangement.
Turn two approved storyboard beats into a connected shot without giving up the intended composition. Matching screen direction and subject identity helps the generated in-between motion stay readable.
Design the ending to reconnect with the opening for a repeatable short-form loop. Similar subject scale, camera position, and background rhythm make the restart less noticeable.
It is an image-to-video workflow that can use an opening image and a finishing image as visual anchors. A prompt describes the action, camera movement, and transition between them. The model creates the in-between motion instead of requiring you to animate every frame manually.
