A Kling prompt for Window Peeper that credits nobody
Updated:
Short answer
Paste the prompt below into Kling and use a photo of an adult or a pet as reference image 1. It lists the zoom, the bite, the stare and the curtain, and asks for a soundtrack without voices so a sound can be added afterwards. It names no poster or song. Expect several attempts, or use Tip Toe AI.

Copy this prompt
Vertical 9:16 image-to-video, roughly 14 seconds. Reference image 1: the CHARACTER, an adult or a pet. Reproduce their face, hair or fur, skin tone, body type and clothing faithfully; a pet sits upright on a chair like a person and holds the food in its front paws. Location: a tall plain apartment block at night, seen from the building opposite. Rows of dark windows, one window glowing warm yellow. Behind that glass: a small kitchen, a little wooden table, a chair and a long brown curtain on the right. Nothing written anywhere. Camera: a neighbour's phone zoom from across the street, slightly grainy night footage, one continuous take. Beats: 1. 0 to 2 s: wide shot of the dark facade with the single lit window, then a fast smooth zoom in until the window fills the frame. 2. 2 to 6 s: the CHARACTER sits at the table holding a plain unbranded cheeseburger in both hands, takes a big bite, freezes mid-chew and stares straight into the lens. 3. 6 to 10 s: still staring without blinking, chewing very slowly, eyebrows low, deadpan and suspicious. 4. 10 to 14 s: puts the burger down, stands, steps to the right side of the window, pulls the curtain shut and glances out once more; the window goes dark. Soundtrack: a light, playful suspense instrumental and faint night-time city ambience. No singing, no lyrics, no speech. Look: photoreal, no brand names, no logos, no on-screen text, no watermark.
Tips for using it
- Set an adult's or a pet's photo as reference image 1, never a child's.
- A written prompt can't pin each beat to a sound, so plan to shift the audio in your editor.
- Keep poster names and brand names out. Filters often block them, so the text only describes the action.
- Tools change quickly. Check which length and aspect ratio yours supports and adjust the opening line.
Questions people ask
Why does the text insist on unbranded food?
Video tools love to invent logos on food packaging, and asking for plain items keeps the clip tidy.
Does it line up with a trending sound?
Only loosely. A text prompt gives the model no timing guide, so beats wander. Tip Toe AI supplies the viral clip's pacing, which is why its videos fit better.
Does this site make me write a prompt?
No. Tip Toe AI uses a preset scene. Add a photo, press the make button, done.
Make a Window Peeper video of your own
Add one photo and you're the one chewing slowly behind the glass. There's no subscription, and the finished video has no watermark.
Make my video