Seasonal dance
Upload 1 to 5 photos to create a 12-second Halloween group dance in 480P, 720P, or 1080P and 9:16, with spooky ensemble choreography built around each subject.
FaceOnTrend generator
Upload one to 5 photos, and each photo becomes one member of the cast. Photos keep the order you upload them in, so select your files in the left-to-right cast order you want on screen. The result is a 12-second 480P / 720P / 1080P 9:16 video.
Before you continue
Prohibited content
Do not upload photos you are not authorized to use. See the full Acceptable Use Policy for details. View Acceptable Use Policy
12s · 200 credits
Aspect ratio
This template is vertical (9:16). 16:9 would require stretching the scene, so it is not offered.
This video costs: 200 creditsEquivalent credit value based on the Creator pack: approximately $5.54Credits are returned if generation fails. View credit pack prices
Generate Once
No account required · Pay first, then upload your photos and confirm authorization · Preview and download in this browser
Failed generations may be retried. Payment refunds are subject to our Refund Policy.
Use Credits
Sign in required · Use your credit balance · Access videos from account history
Need more credits? See pricing
Powered by Genjutsu 1.0 · Model info
Halloween Dance is the collection's ensemble format. Upload between one and five subjects and the template arranges them into a 12-second spooky group performance, preserving the supplied image order as the cast grows. The 9:16 result is available in 480P, 720P, or 1080P and combines seasonal styling with synchronized choreography, working naturally for Halloween posts, costume reveals, friend-group edits, party invitations, TikTok, Reels, and Shorts.
Each source must be a JPG, PNG, or WEBP file no larger than 5 MB. Choose clear, well-lit images with the subject's important identifying features visible. For a choreography-heavy template, photos that show more of each subject's overall appearance can give the model useful context, but they do not guarantee exact footwork or anatomy in every frame. Keep the intended left-to-right character order in mind when selecting multiple files. This format differs from the single-subject dance pages by coordinating a group, so consistency across lighting, crop, and image quality can help the cast read as one scene.
Halloween group dances are a seasonal short-form format: a cast in costume performs the same synchronized choreography together in a single vertical frame. The recognizable appeal is the group sync itself. When every performer hits the same steps at the same time inside one spooky-but-playful scene, the result reads instantly in a feed, even before sound is turned on. That is why the format keeps reappearing every October across TikTok, Reels, and Shorts, and why group-sync dance videos continue to be remade with new casts, new costumes, and new AI tools each season.
The template reproduces that structure with AI. One private reference video guides the choreography, the timing, and the scene, while your uploaded photos supply the cast. The model creates a new performance around your subjects rather than pasting their faces onto the reference performers, so the group, the costumes, and the details come from your uploads instead of from someone else's footage. Because every run is a fresh creation, results vary between generations.
The format suits costume reveals, friend-group edits, party invitations, character showcases, and any Halloween content where a coordinated group reads better than a solo shot. The vertical 9:16 frame holds a full lineup of dancers, which is exactly the composition short-form feeds favor.
The Halloween Dance template is built around a 12-second, reference-led group performance. The reference video supplies movement timing, camera guidance, and the choreography for a Halloween-themed dance, while your uploaded photos determine who performs it. You can watch the demo to see the source performance; exact cuts, individual gestures, and moment-by-moment timing are not guaranteed in a generated result.
The goal is a coordinated dance with one to five uploaded subjects. Each subject's movement is adapted to its visible appearance and, where applicable, its species and anatomy. The reference helps guide the performance without fixing the exact sequence of shots or promising identical choreography in every output. Positions, gestures, and scene details can vary between generations.
The number of photos you upload changes the size of the cast, not the duration or the resolution.
Upload between one and five photos. Every file must be a JPG, PNG, or WEBP image no larger than 5 MB, and each photo becomes one member of the cast. Choose clear, well-lit images where the subject's important identifying features are unobstructed, and avoid heavy occlusion, strong blur, or accessories that hide the face. Because the choreography involves the whole body, photos that show more of each subject's overall appearance, ideally a full-body or three-quarter view, give the model more useful context than a tight face crop.
Upload order matters. The template preserves the order in which you supply the images as the cast grows, so select your files in the left-to-right cast order you want to see on screen. Keeping lighting, framing, and image quality consistent across all photos helps the group read as one scene rather than a mix of unrelated shots.
Costumes in the photo are the costumes in the video: the model recreates the scene from what your uploads show. Clear, well-lit costume details survive the generation better than dark or cluttered ones, though no image guarantees that every outfit detail or dance movement is reproduced exactly. The same guidance applies to pets, toys, and illustrated characters, whose movements are adapted to their own anatomy instead of being turned into humans.
You can upload between one and five photos, and each photo becomes one member of the cast. A single image leads the scene, while several images create a fuller group composition. The configured duration stays 12 seconds regardless of how many subjects you include.
Yes. With one upload the subject leads the scene on their own, and the same synchronized choreography, 12-second duration, and vertical format still apply.
Yes. The template preserves the order in which you supply the images as the cast grows, so select and upload your files in the left-to-right cast order you want to see on screen.
Choose photos where the costume, its colors, and the subject's key features are clearly visible and well lit. The model recreates the scene from what the photo shows, though exact costume details are not guaranteed in every frame.
Yes. People, pets, toys, and illustrated characters are all supported. Dance movements are adapted to each subject's species and anatomy, so a pet stays a pet instead of being turned into a person.
The finished video is 12 seconds in a 9:16 vertical frame, available in 480P, 720P, or 1080P.
The reference video guides one synchronized group choreography for the whole cast, so all subjects follow the same sequence. Because each generation is a new creation around your uploads, individual positions and movements can vary between runs and are not reproduced frame by frame from the reference.
Charged credits are returned to your balance when a generation definitively fails. On templates that offer the One-Time Pass, failed generations may be retried instead, and payment refunds are subject to the Refund Policy.