Crazy Frog AI Prompt: Copy-Paste Prompts for the Trend

Format: A 12-second clip at 720p, 9:16. No speech, no captions and no lettering anywhere in the frame.

Reference: Use only the adult from the attached photo and keep their face, hair and build recognizable. Do not invent a second rider. The rider stays the sharpest thing on screen.

Clothes: Keep exactly the outfit from the photo, same colors and shape, with no added logos.

Other people: Nobody else on foot. One plain silver car without badges, logos or readable plates chases close behind; the driver is only a shadow behind the windscreen.

Location: A bending overpass at dusk drawn like a cutscene from a very old video game, a bit grainy, with round towers and boxy apartment blocks; or a sunny city street, an open highway or a wet neon street at night. No readable signs anywhere.

Framing: Quick cuts about once a second: low front shots at road level, a high wide shot along the bend, side shots and a close-up of the car’s wheel. Every time the rider is on screen, show the whole body and the empty gap underneath.

Movement: 0–3s: Hang in the air just above the asphalt in a low sitting squat, fists out on handlebars nobody can see, racing at a low camera as a silver car closes in, then roll the right wrist like opening the throttle; 3–6s: Slide along the road in that same floating squat, shown from high above the bend and then from the side, hair and clothes blown back while the car catches up; 6–9s: A quick shot of the car’s wheel flashes by the lens, then back to the rider, still hovering, who peeks over a shoulder at the car and grins; 9–12s: Hit a bump, flip into a goofy roll that does no harm, sit up laughing with fists still up, and watch the car swerve by and drive away.

Audio: Soft background noise only, like wind, a humming scooter and the car’s engine, with no music or voices.

Avoid: Watermarks, changing facial features, extra fingers or limbs, visible vehicles under the rider, injuries, text overlays, logos and recognizable public figures.

This is a paraphrase. It differs from the prompt used by this site.

Format: A 12-second clip at 720p, 9:16. No speech, no captions and no lettering anywhere in the frame.

Reference: Two separate adults, one per attached photo. The person from photo 1 rides on the left and the person from photo 2 on the right for the whole clip; never blend or swap their faces. Both stay sharper than the background.

Clothes: Each friend keeps the outfit from their own photo, same colors and shape, with no added logos.

Other people: Nobody else on foot. One plain silver car without badges, logos or readable plates chases close behind both riders; the driver is only a shadow behind the windscreen.

Location: A bending overpass at dusk drawn like a cutscene from a very old video game, a bit grainy, with round towers and boxy apartment blocks; or a sunny city street, an open highway or a wet neon street at night. No readable signs anywhere.

Framing: Quick cuts about once a second: low front shots at road level, a high wide shot along the bend, side shots and a close-up of the car’s wheel. Keep both full bodies, the empty space beneath them and a clear gap between the two riders visible whenever they are on screen.

Movement: 0–3s: Both riders hang just above the asphalt in low sitting squats, each gripping handlebars nobody can see, racing at a low camera as a silver car closes in, then both roll their right wrists like opening the throttle; 3–6s: The pair slide along the road side by side, shown from high above the bend and then from the side, clothes flapping as the car catches up; 6–9s: A quick shot of the car’s wheel flashes by the lens, then back to the two riders, still hovering, who peek back at the car and laugh; 9–12s: Both hit a bump, flip into goofy rolls that do no harm, sit up laughing, and watch the car swerve by and drive away.

Audio: Soft background noise only, like wind, a humming scooter and the car’s engine, with no music or voices.

Avoid: Watermarks, changing facial features, merged or swapped faces, extra fingers or limbs, visible vehicles under the riders, injuries, text overlays, logos and recognizable public figures.

This is a paraphrase. It differs from the prompt used by this site.

How to use these prompts

Paste a prompt into an image-to-video tool that accepts reference photos and attach your photo. Results vary by tool: some keep the face better, some add a moped anyway. If a vehicle appears, repeat that nothing is under the rider. These prompts are free to copy, but the tools you paste them into may charge; compare routes in the free guide. On this site you skip the prompt entirely.