You have a photo and an idea for a short video. The next decision is what you want to choose yourself: a ready-made scene, the movement from a particular clip, or a change to a face in existing footage.
Those are different starting points. A page that says “motion transfer” does not necessarily accept your own video, and a template that makes a pet dance does not necessarily offer a different dance, soundtrack or camera move.
Genjutsu Lite currently offers three prepared photo-to-video templates using Seedance. It is an independent tool, not Higgsfield's official Genjutsu Lite or Light. Uploading your own reference video here is still coming soon.
Start with the part you need to control
| Your goal | Workflow to look for | Check before submitting |
|---|---|---|
| Put your photos into an existing pet dance, solo dance or car scene | A prepared video template | The complete sample, photo slots, fixed format and sound |
| Use the movement or camera sequence from a particular clip you have | A tool that explicitly accepts a motion-reference video | Whether uploads are open, permitted clip length and which parts the tool may change |
| Change a face while keeping an existing performance | A face-focused editing workflow | What happens to the body, clothes, background and audio; do not infer this from the label alone |
These are ways to choose a workflow, not guarantees about how precisely a model will follow your inputs. Products can combine several modes, and the controls available in one mode may be different from another.
A prepared template chooses the scene for you
A template is useful when its existing result is already close to your idea. You supply the required photos, while the tool provides the prepared routine or scene.
For example, if you want two pets to perform the NEON routine, begin with the complete duo-dance example. If the idea is a solo performance or a fictional car stunt, watch Beat Drop or Roadster Ramp.
Watch past the first few seconds. Check whether the camera, ending, orientation and sound suit what you want to share. At Genjutsu Lite, these are prepared scenes: selecting a new photo does not let you upload different choreography or choose another song. The existing sample also does not turn into a personalized preview when you select a photo.
For the exact requirements, use the guide on the matching template page rather than assuming that any portrait or group photo will fit.
A specific reference clip needs an explicit upload workflow
Suppose you want to follow a particular performance you filmed. Finding a similar-looking template is different from supplying that clip as the motion reference.
Look for a tool that explicitly accepts reference-video uploads in the selected mode. Check the allowed length, image count and whether the original timing, camera, subject or setting is intended to change. You must have permission to use the reference as well as the photos.
As one product-specific example, Higgsfield's release notes describe its official Genjutsu Lite release of 19 October 2026 as accepting a 4–15-second video and up to ten reference images, with Motion transfer and Objects swap modes. Those are Higgsfield's stated inputs, checked again on 30 October; they do not describe this site's Seedance templates or establish a result we have tested.
Face swap is not a promise of new choreography
If your goal is mainly to change a face in an existing performance, inspect a face-focused tool's examples and controls. Ask what it changes outside the face, whether the original performer remains recognizable, and how it handles turns or obstruction.
Do not assume that face editing will recreate a full body, change an outfit or preserve every frame. Equally, a full-body motion-transfer tool should not be treated as a precise face-only editor. Use authorized images of consenting adults and check the service's rules before submitting.
Compare the quote units, not just the credit number
A rate per second and a price per generated clip cannot be compared by the number alone.
When checked on 30 October 2026, genjutsu.online displayed a per-second credit rate and a model multiplier for its selected Kling motion-control option. Its page described other pricing paths as well. This is an observation of the public interface; we did not verify checkout, the underlying model or a generated result.
Genjutsu Lite currently quotes 100 credits for each of its three available templates. For a per-second quote elsewhere, check the billable duration and any applicable model or resolution multiplier. Then compare the actual scene, length, sound and output settings. A larger credit balance or a lower headline rate does not by itself mean a cheaper usable video.
Before you choose a tool
- Decide whether you want a prepared scene, your own reference movement or a face-focused edit.
- Confirm that the selected mode accepts the inputs you have permission to use.
- Watch a complete result and read its format, sound and effect limitations.
- Check the final quote and what happens if the task fails or its outcome is uncertain.
For the current Genjutsu Lite options, start with the three real examples. Selected photos stay in your browser until you explicitly start generation. Signing in, trying example photos or buying credits does not automatically generate a video.
The examples show individual existing results. They are not a success-rate measurement or a guarantee of identity, motion or completion time for your photos. If you already submitted a task, see task status, credits and download deadlines rather than submitting it again because you are still waiting.