Product hero with a scripted ending
Break a product reveal into a wide shot, a detail shot, and a closing composition. Review labels, shape, and lighting in each clip before deciding which ones fit your edit.
Plan your sequence shot-by-shot. Describe each action and use a supported reference to guide framing. Generate a separate clip per request for your external timeline, then review the result and download the clips you select.
Plan a few shots at a time. Each request produces one clip, so your shot list becomes a practical guide for what to generate next.
Break a product reveal into a wide shot, a detail shot, and a closing composition. Review labels, shape, and lighting in each clip before deciding which ones fit your edit.
Use the VELVET DROP poster to set a starting composition. Plan one camera move per request, then check lettering, fabric, and subject changes in the returned clip.
Use the vase-and-books still life to plan a restrained interior shot. Describe the light, framing, and one intended movement, then check object shapes before using the clip.
Use the four-panel material board to describe texture and color choices in your shot list. Generate each scene separately and compare; the board guides the brief but does not ensure continuity.
Each generation request is a chance to test one idea. Keep the goal concrete, note what worked, and give your editor a clip that fits the cut.
State what the clip must do for the sequence: establish a place, show a product, or follow a person through one action. A specific goal makes the returned clip easier to judge.
If the selected model accepts a reference, choose an image that shows your intended framing or subject. Confirm the available input types before building your request around one.
Watch the beginning, middle, and end of the returned clip. Notice subject changes, unwanted motion, and whether the action leaves a clean point for the next cut.
Download the clips you want to keep and arrange them in your video editor. Add pacing, transitions, text, and sound there; this workspace does not include a timeline editor.
These are example briefs to begin clips for a sequence. They describe a shot and ask you to check motion, not show a generated result.
A wide shot of a busy coffee shop. The camera slowly moves forward through the door. People are sitting at tables. Steam rises from cups. Warm lighting.
Does the camera movement create a sense of entry? Is the steam visible?
A close-up of a barista pouring milk into a coffee cup. The camera is static. The milk creates a pattern. The barista's hands are steady.
Is the pouring motion smooth? Is the pattern in the milk visible?
A medium shot of a customer taking the coffee cup from the barista. The camera pans slightly to follow the hand movement. The customer smiles.
Does the pan follow the hand movement? Is the customer's expression clear?
A wide shot of the customer walking out of the coffee shop door. The camera holds on the door as it closes. The street outside is visible. Daylight.
Does the door close naturally? Is the street scene consistent with the interior?
Generate specific visual assets that integrate into your production pipeline. Focus on creating usable clips that match your creative direction and are ready for external editing.
Prioritize the missing shot in your sequence—for example, a close-up of an action—before creating variants of the same shot. This builds a useful shot list efficiently.
Sign in to view the credits currently available to your account. An existing balance is required for generation while new purchases remain paused.
Model and output choices affect the displayed credit cost. Read the estimate for your current request instead of assuming it matches another model or previous attempt.
Give each request a specific evaluation goal. Assess the result before deciding if another attempt is valuable, and keep notes on the brief you want to revisit.
If a request fails or its status is unclear, check the task status and your account balance before resubmitting. Contact support with the task reference if the outcome needs investigation.
Your text prompt and a supported visual reference provide distinct information to the generation. The prompt describes the action, mood, and sequence. The reference offers composition, color, or subject style. The model interprets both to produce a clip. Final layout, typography, and editing happen in your external software.
Write your prompt as a clear brief for a single shot. Specify the subject, action, environment, and desired camera perspective.
Upload an image to influence the visual style, composition, or subject appearance. The model uses this as a guide, not a blueprint.
Combine your text and reference to steer the shot's framing and movement. Instructions in your prompt can request specific layouts or motion.
Download a clip that serves as a raw asset for your project. Edit, sequence, and add typography in your preferred timeline software.
Answers about the site identity, request workflow, references, credits, and the nature of the concept examples.
Understand the distinction between the website brand and the actual model selected for a generation request.
Learn about supported inputs and how to evaluate the returned details before using an output.
Use separate requests for your planned assets and complete final assembly or detailed editing in your external tools.
Understand the paused purchase flow and the purpose of the concept illustrations.
Write separate briefs for each clip describing subject, action, and camera movement. The model generates individual clips for external assembly.
Choose a supported model, write one clear brief, and evaluate the result against its intended use. Check your account credits before submitting a generation request.