A window scene with the setup already chosen
The Caught at the Window template is a fixed motion preset for the trend of the same name. Instead of writing a prompt or assembling an edit, you provide one suitable photo and the scene does the rest: a dark apartment block at night, one warm window, and someone caught enjoying a burger they clearly did not expect to share with an audience.
Here, "template" means a fixed generated scene. It is not a downloadable editing project and it is not your face pasted onto someone else's footage. A new video is generated for every request: the reference clip guides the camera and the action, while your photo supplies the person.
Watch the demo on this page before you spend credits. If you are ready to create, go straight to the Caught at the Window generator. For where the format came from, read the trend page.
The four beats that make the joke
The push-in
The clip opens on the building. Most windows are dark; one is lit. The camera glides toward it like a passer-by who has just noticed something. This slow approach is what makes the payoff work: the viewer arrives at the window a half-second before the person inside looks up.
The mid-bite moment
Inside, the subject sits at a small table by the window with a burger in both hands, a paper takeaway bag and a burger box next to the plate. They are mid-bite, unguarded, lit by warm indoor light against the blue night outside. Keep your source photo clear and front-facing; this is the shot your face has to carry.
The stare
Halfway through, they notice the camera. The bite freezes. The expression does the comedy: not fear, just the flat annoyance of someone whose private snack has become public. The generator aims for that beat in every render, and a photo with visible eyes and an unobstructed face gives it the best chance.
The curtain close
They set the burger down and pull the curtain shut. The last frames are the closed curtain with warm light leaking around the edges. It is a clean, loopable ending, which is one reason the format travels so well on TikTok and Reels.
What stays fixed and what can vary
The building, the window, the table setting, the camera move, the timing and the lighting are fixed by the preset. Your photo determines who is sitting at the table: face, hair, build and outfit all come from the image you upload.
Generation is still AI, so small details vary: hand positions, the exact expression, background objects and how the curtain falls. Treat the demo as the scene's promise, not a frame-by-frame guarantee.
Template or CapCut?
Both routes exist for this trend. A CapCut template keeps you inside the CapCut app and swaps a clip into a pre-cut project. This generator works in the browser, builds the scene around your photo with an AI video model, and hands you a finished file you can post anywhere. The CapCut workflow guide compares the two routes honestly, including when the free route is the better choice.
Ready when you are: upload a photo in the generator, or check the credit packs first.