Start with a still image and describe how it should move. Choose a video model, review the settings, and create a clip around one clear action.
Start frame
01Switch models without leaving the shared workspace.
02See the final credit cost before every generation.
03Find finished work in your results and history.
Image-to-video examples
See what one frame can become
Character motion · HappyHorse 1.1A paper fox takes a short walk through a layered paper forest.Camera movement · Kling 3.0A simple side profile supports clear, readable movement.Product reveal · Seedance 2.0A compact camping stove becomes the focus of a short product reveal.
Starting imageThe original still defines the front view, materials, lighting, and surrounding space.Kling 3.0 outputThe changing viewpoint reveals the side and rear of the object. Compare its shape and details with the starting image.
Creative guide
Prepare a starting image that can support motion
A strong image-to-video result starts with a frame that already communicates subject, composition, and depth. This exact pair lets you compare the source image with the motion added to it.
Use a readable source
Choose a sharp image with a clear subject and enough room for the planned movement. Awkward crops, tiny details, and overlapping limbs often become harder to control once motion begins.
Leave room for movement
Give a turning subject space beside it and a push-in shot depth behind it. The source composition should support the motion rather than force the model to invent missing space.
Protect essential details
Keep critical text, identity, and product geometry large and visible. Generated movement can change hidden surfaces or fine details, so review the complete clip before approval.
Check the first frame
Open the source at full size before uploading it. Compression artifacts, accidental text, weak edges, or an already blurred subject can become more noticeable once the clip introduces movement and changing detail.
Motion prompt to try
“Make one slow clockwise orbit around the field recorder. Keep the device fixed on the platform and preserve its shape. Use gentle background parallax and finish on a steady three-quarter view.”
Image-to-video outputA selected output illustrating a focused motion instruction.
Creative guide
Describe what changes between frames
The image describes appearance; the prompt describes change. Keep the motion instruction focused enough that you can identify whether the source, the action, or the camera caused an unwanted result.
Separate subject and camera
Write the subject action first, then the camera move. A coat fluttering while the camera slowly pushes in is easier to interpret than a vague request for cinematic motion.
Use one motion goal
Begin with one dominant change: a turn, orbit, reveal, or environmental movement. Add another action only when it supports the same shot and remains easy to evaluate.
Define pace and ending
Use words such as slow, steady, or quick with a visible action, and describe the final frame when it matters. Avoid stacking several mood adjectives that do not change movement.
Keep appearance words selective
Mention light or atmosphere only when it should visibly change during the shot. The source image already carries much of the look, so motion instructions should keep priority over repeated style description.
Motion prompt to try
“Use one deliberate camera move and one subject action. Keep the background stable, preserve the main silhouette, avoid scene cuts, and end on a frame that can hold for editing.”
Creative guide
Choose a small scene with a clear purpose
Image to video works best when the still already carries the visual idea and motion adds a clear purpose. These are practical ways to keep a short clip useful.
Product reveal
Start from a clean product view and use a restrained orbit, turntable motion, or light change. Watch labels, edges, and proportions throughout the clip.
Portrait moment
Use a stable pose and ask for one subtle expression, head turn, or fabric movement. Small motion is often easier to approve than a dramatic change in stance.
Landscape atmosphere
Let clouds, mist, water, foliage, or light provide movement while the composition stays recognizable. Avoid combining a large camera move with several environmental changes at once.
Editorial transition
Plan an ending that can hold, cut, or connect to another shot. Check both the opening and closing frame instead of judging only the most attractive moment.
Loopable detail
Choose repeating motion such as steam, fabric, water, or a mechanical cycle, and keep the camera stable. Compare the opening and ending so the clip can be trimmed into a smoother loop.
Social cutaway
Use a source with a strong subject at small viewing sizes. Ask for one immediate action, preserve safe space for later captions, and judge whether the idea reads without relying on sound.
Creative guide
Choose a video model around the shot
Choose a model by the inputs and settings shown in the generator. The best choice is the one that supports your source image, shot length, final placement, and the detail that matters most. Start with the option whose visible controls match the plan, then compare another model only when a specific weakness remains. Keep the source image and main motion instruction the same so the comparison gives you a clear answer.
Input fit
Confirm that the selected workflow accepts the source you have and exposes the reference controls you need. Do not assume every model uses the same input arrangement.
Shot control
Compare the available duration, aspect ratio, resolution, and audio choices with the planned shot. Controls that do not affect the task should not drive the decision.
Review burden
Detailed products, recognizable people, and exact typography need closer continuity checks than atmospheric scenery. Choose a simpler shot when approval accuracy matters more than spectacle.
Visible cost
Read the credit amount shown before submission and recheck it after changing settings. If generation is unavailable, keep the source and prompt ready rather than inventing unsupported settings.
Iteration value
A model is useful when its result gives you a clear next decision. Compare one controlled prompt across suitable options when needed, then continue with the choice that best preserves the parts your project cannot trade away.
Get started
From first idea to a useful result
01
Add the starting image
Open image to video and choose a clear source image. Check its crop, the space around the subject, and any details that must remain recognizable. You can prepare a new image in the image generator first if you do not yet have a suitable starting frame. Read the upload requirements before selecting a file.
02
Describe the motion and select settings
Write the main action, camera intention, pace, and ending. Choose a video model, then review its current duration, resolution, framing, and audio controls. Check the credit amount before submission. If the generator reports that a requirement is missing or generation is unavailable, follow that message; adding extra prompt text will not resolve an account or availability issue.
03
Review, download, and refine
Play the whole video, including the last second. Check for drifting features, abrupt cuts, changing backgrounds, and unexpected movement. Download a useful version before making another attempt. If the clip misses the brief, decide whether the next change belongs in the source image, motion instruction, or settings, then revise that part deliberately.
Plans for image and video
Pricing
Annual plans are billed upfront. The same plan credits refresh every 30 days and expire at the end of that window; unused subscription credits do not roll over.
Free
$0/ month
Explore the workspace and prepare a first request before choosing a paid plan.
The image shows what the scene looks like, but it does not fully specify what should happen. Use the motion prompt required by this workspace to describe the action and camera intention. A brief, focused instruction is more useful than repeating every object already visible in the source.
02Can I use several starting images?+
Input requirements depend on the selected model and workflow. Follow the reference controls shown by the generator rather than assuming that a gallery of images becomes a sequence. This page describes animating a starting image, not a slide show builder or a timeline with independently timed scenes.
03Will the video exactly preserve my photograph?+
Generated motion can change details, particularly when the subject turns or the camera reveals hidden surfaces. Use a clear source, keep the motion focused, and review the full result. When exact text, identity, or product geometry is essential, plan for a careful approval or editing step.
04Can the generated video include sound?+
Some model selections expose an audio option. Check the controls for the model you choose instead of assuming sound is available everywhere. If you request audio, listen to the complete result and judge its relationship to the scene. You can also plan to add a soundtrack separately in a video editor.
05What if the motion is too strong?+
Reduce the requested action or camera movement and describe a slower, more limited change. Keep other instructions stable so you can judge the difference. If the source image suggests an awkward pose or an unstable composition, revising the image may be more effective than adding more motion constraints.
06What if I do not have a suitable image?+
Create or prepare one before starting the video. The image workspace can help explore a composition, while text to video is an alternative when you want the video model to establish the scene itself. Choose the path based on whether a particular starting appearance is important to the result.
Your next step
Start with a frame worth moving
Choose a readable image and one clear motion goal, then review the clip from beginning to end.
