
Good starting point
- One clearly visible face in each image
- Front-facing or slightly turned, with hair visible
- Even light without strong filters
- A portrait you are comfortable sending for processing
Two photos for the little knight and their protector.Two photos, one gentle sword-to-toy story. Cast a little knight and their protector in the misty-field scene.
A heavy sword becomes a soft toy. The small knight no longer has to carry it alone.
The shown scene follows a small knight on a misty field. A taller protector arrives, kneels, takes the sword, and gives a plush rabbit. The ending holds both characters together.
View the red-cape creator's reelYour photos provide facial identity and hair. This template keeps the reference armor, capes, character sizes, setting, and props. It does not copy your everyday clothes.
Knight sword swap, sword-to-toy, and inner child knight describe related versions of the same handover idea.
The reference on this page is the red-cape scene shared by @gazizaaz. Watch it before assigning your two photos.
Watch @gazizaaz's reel@ray_chieee describes another knight reel as playing both her younger and adult self. That is a storytelling option, not a requirement to upload a child's photo.
View @ray_chieee's versionUse two clear portraits for a fictional cast. The template fixes the scene and handover, so no story prompt or gift selection is needed.
Match each photo to its role, check the example, then review the complete handover.
Place the small knight's portrait first and the protector's portrait second. Use one clear face per photo.
The complete story is 15 seconds. Choose 480P or 720P and review the total credits before you start.
Watch both faces, the sword handover, and the toy at the end. Download the finished MP4 or return to it in History.
Face and hair detail matter here. The armor and small-versus-adult proportions come from the scene.


The roles and props are prepared. Focus on choosing the two faces you want in the scene.
The first portrait belongs to the little knight; the second belongs to the protector, including when they kneel together.
The prompt keeps the source armor, capes, and body scale while requesting each photo's face and hair.
The 15-second reference includes the approach, kneeling, sword transfer, toy, and protective ending.
Hands, props, and likeness may drift during contact. Watch the full result before sharing it.
One generation covers the 15-second story. Resolution changes the total credits.
480P is the default. Check the current price on the generation button after choosing a resolution.
Use your existing credits. A new attempt is a separate generation and has its own displayed cost.
The template uses Wan3 Prime through fal. Technical failures return credits; a completed video with visual imperfections is not a technical failure. View the model documentation
Check the full action, not only the opening and ending frames. Different templates can tell different knight stories.
Keep using the same account and credits across the template library.
It is a short AI story about a small knight with a heavy sword and a protector who takes the sword and gives a toy. This page uses the red-cape scene shown in the reference, not a battle or a freely written story.
Photo 1 is the little knight holding the sword at the start. Photo 2 is the adult protector who arrives and kneels. Both photos are required; this version does not support replacing only one role.
No. This template uses the photos for face and hair, while retaining the reference armor, capes, and small-versus-adult body proportions. An adult portrait in the little-knight slot creates a fictional small knight, not a reconstruction of that person's childhood.
Not in this version. The sword, plush rabbit, misty field, and handover sequence are fixed to the reference. There is no extra object photo or free-text story input.
The complete clip is 15 seconds in the source's 4:3 frame, with 480P or 720P output. Keeping the original frame leaves room for the sword, both characters, and the toy.
No perfect match is guaranteed. New frames are generated, and fingers, the sword grip, toy contact, and faces can shift. Review the kneeling and handover closely before sharing.
Yes, as fictional casting. You do not need to upload a child's photo. If you choose a childhood photo, use your own or one you are authorized to use, avoid identifying details such as school badges, and consider whether you want the result shared publicly.
Technical failures return the generation credits automatically. Visual imperfections in a completed result are not a technical failure. Starting another attempt uses additional credits.
No prompt is needed. The two-role prompt and reference are configured. The finished video uses the prepared reference audio; display previews are muted. Rights to footage, music, and portraits are separate from the generation service.
Delete a completed entry from account History. This does not instantly erase copies held by generation or storage providers. See the privacy policy for data-handling details.