One photo puts you in the Window Peeper Filter scene: eating a burger by a lit window, noticing the camera, then closing the curtain. Watch the matching 14-second reference and make your own vertical clip.
One portrait · 14 seconds · 9:16
The dedicated window clip supplies the bite, stare, and curtain movement. The model generates a new scene from your portrait; results can vary.
One photo and permission required. Review credits before generating.
Supplied Window Peeper reference · 14s · not generated by this site
Still from supplied reference · not a generated result
The opening establishes a dark exterior and a warm window, with the subject seated at the table eating. This still comes from the supplied reference clip. It shows the planned setup, not a personalized result or a sample generated by this site.
Back to this generator
Still from supplied reference · not a generated result
The pause is the main comic beat. The person notices the camera and looks outward while holding the food. A clear portrait gives the model more facial detail to work with, but eye direction and expression are generated and can differ from this reference still.
Back to this generator
Still from supplied reference · not a generated result
The final movement brings the curtain across the window. Keeping the full 14-second template gives the model room for that ending. Watch the hands and fabric in the result; a distorted curtain or a missed closing action can occur even when generation completes successfully.
Back to this generatorOne portrait · 14 seconds · 9:16
The Window Peeper Filter is a staged short-video joke built around a simple sequence. A camera watches a lit apartment window at night. Inside, someone is eating a burger, notices the viewer, holds an awkward stare, and eventually closes the curtain. Here you supply one portrait and the generator asks for a new version with that person in the scene. This is a photo-to-video template, rather than a live camera effect. The preview above uses the supplied reference footage, so you can watch the exact scene being used to guide the motion.
Choose a JPG, PNG, or WebP image up to 10 MB. A front-facing subject with visible shoulders and even lighting works best. Only one photo is needed: this scene has one person at the table. Avoid group pictures or a hand covering the face.
Try this step in the generatorUpload your photo references
Enlarge interface screenshotThe template stays at 14 seconds, vertical 9:16, using Wan 3.0 and its own motion reference. Choose 480p, 720p, or 1080p. If reference audio is offered, decide whether to include it. The credit total updates when you change quality.
Try this step in the generatorChoose the video settings
Enlarge interface screenshotSign in, confirm your input permissions, and submit with sufficient credits. Watch the progress display and wait for the finished preview. Check that the same person stays visible throughout the bite, stare, and curtain movement. Download the MP4 after reviewing the result.
Try this step in the generatorConfirm and create
Enlarge interface screenshotThis effect uses a separate 14-second reference, not the orange-booth rap routine elsewhere on the site. The reference guides the camera approaching the window, the person handling food, the stare, and the curtain closing. Your portrait supplies the new subject. The model is instructed to avoid transferring the reference actor, packaging logos, and text into the result. Because it generates new frames, it may still alter the face, food, hands, or curtain. Treat the reference as the target choreography rather than proof that your own result will match every detail.
Back to this generatorCreator posts from the Window Peeper trend. Videos load as you scroll to them and start muted when your browser allows autoplay. Open the original post for sound or if TikTok is unavailable.
Use a sharp image with one subject looking toward the camera, preferably with some shoulders or upper body visible. Heavy sunglasses, deep shadow, extreme side angles, and faces cropped at the chin make the reference less useful. You do not need to hold a burger or stand beside a window in the input photo; those belong to the generated scene. If the source image only shows a head, the model must infer the rest of the outfit and body, so expect less control there.
The order matters more than any caption: the camera approaches, the subject eats, the subject notices, and the curtain closes. This page keeps the motion source, duration, and vertical frame fixed so a stray rap preset cannot replace that sequence. There is no second performer, no microphone, and no topic for writing lyrics. Choose another tool from the footer if you want a music performance. For this effect, start by watching the reference once all the way through, then compare those same beats in the finished clip.
The reference preview includes its source sound. Your render only receives a separate reference track when the site offers that option and you leave it selected. Without one, the prompt asks for quiet room ambience rather than a rap song. A model can still change audio, so listen before posting. The MP4 is suitable for a vertical editing workflow; you can add permitted captions or licensed music in your own editor afterward. This page does not publish automatically to TikTok or attach an official platform filter.
The generator shows a quote for the current settings, and the server recalculates it before submitting a job. Credit packs are one-time purchases. Watching a reference clip or choosing settings is not a personalized free render. Credits are deducted when a generation starts; a technical failure returns them automatically. A successful but imperfect result is charged. Read the pricing page for current packs and the terms for purchase rules.
One-time credit packs. A 14-second Window Peeper Filter video at 480p uses 13 credits. Packs never expire.
Single
$9.99
one-time
$12.99 / 14s at 480p
Duo Pack
$29
one-time
$7.54 / 14s at 480p
Party Pack
$59
one-time
$3.83 / 14s at 480p
Crew Pack
$99
one-time
$3.68 / 14s at 480p
Studio Pack
$199
one-time
$3.23 / 14s at 480p
It is the staged window joke: someone eats a burger in a lit room, notices the camera outside, pauses to stare, then closes the curtain. This page uses the supplied 14-second reference to guide that sequence and your portrait to cast the person. The initial preview shows the reference, not a personalized result.
Add one clear portrait, choose 480p, 720p, or 1080p, and review the credit total. Keep the built-in reference sound selected if that option is available and you want it. Confirm your permissions and generate. The template already includes the window, burger, camera movement, and curtain, so you do not need photos of those objects.
This is a separate browser tool that renders a new video. It does not open TikTok’s camera or install a CapCut template. The creator clips on this page are external references to the format. Their playback and any associated templates are controlled by the original creators and platforms, while this generator uses its own configured reference clip.
This version has one person at the table and accepts one portrait. Use a photo with one clear subject; a group picture does not add another character to the scene. If two friends want to try it, each can make a separate clip. The two-person rap pages offer a different workflow for sharing a performance.
The current template is the burger scene, fixed at 14 seconds and vertical 9:16 on Wan 3.0. Choose output quality, but keep the sequence intact so the bite, stare, and closing curtain all have time to happen. Alternate foods, uploaded motion clips, different durations, and Seedance are not offered for this version.
The reference preview has its source audio. A separate built-in track is sent to your render only when the site offers it and you keep it selected. Without that track, the prompt requests quiet room ambience. Listen to the result before posting, since AI can change sound; music permissions remain your responsibility.
The reference is free to watch, but it is not generated from your portrait. A personalized render uses the quoted credits for the fixed 14-second video and its motion reference. Quality changes that total, and the server recalculates it before submission. You can use a one-time credit pack; no subscription is required.
Check the whole result for identity changes, hands, the burger, and the final curtain movement. A clearer portrait may help on another attempt. Technical failures return credits automatically, but a successful render can still be imperfect and is charged. Download the completed MP4 from the preview, or reopen the job in My videos.
Inputs and completed videos are stored privately; they are not automatically added to a public gallery. The initial preview and external posts are references, while scenario illustrations describe ideas. AI may change faces, hands, timing, or sound. Use only photos, footage, and recordings you have permission to use. This independent tool is not affiliated with TikTok, CapCut, the original performers, or template creators. Download a local copy of a result you want to keep, and review it before publishing.
Start from the main studio: bring two portraits together, explore three stages, and choose a reference track or an original rap.
Explore this video
Make a performance for two friends or a couple. Choose who leads, or switch to an original rap and trade lines.
Explore this video
Recreate the orange-booth routine with two portraits, one hanging microphone, and a lead-and-hype performance.
Explore this video