P—20 / TOKYO FISHEYE STREETWEAR

Y2K fisheye streetwear prompt for MiniMax H3

Thirteen seconds, six locations, one ugly lens — and a zipper that is the first cut.

A fashion-video prompt that is really a camera prompt. The clothes and the two men are specific, but what holds six locations together is an 8mm fisheye that never goes away, plus two transitions that are physical objects: a zipper opening, a palm closing over the lens.

Steps
One shot
Mode
Text to Video
Duration
13s
Ratio
16:9
Resolution
2K
Audio
Boom-bap + street tone
Cost
$1.69

Reference render — not generated on this site. Source: inspova

Output reference

Two streetwear kids under Tokyo neon, shot on a raw fisheyeVideo
The clip
2980 / 7000

Images 0/9 · Clips 0/3 · Audio 0/3 · 12 files max

Sign in free · 3 clips a day

No card, no account: watch and download 12 real MiniMax H3 clips, and run one prompt of your own on Agnes Video V2.0 — 16:9, watermarked, once. After that, sign in free for 3 clips a day on the free engine. MiniMax H3 on your own prompt starts at $24.9 a month.

1 FREE CLIP · WATERMARKED

01 — Inside

One prompt, one workflow

The clip beside the verbatim source. Copy it into the console above, change what you need, generate.

Reference output

Two streetwear kids under Tokyo neon, shot on a raw fisheyeVideo
The clip
The promptthe clipMiniMax H3 · Text to video
2,980 chars

The night

Timed to the second. The zipper and the palm are the only cuts the prompt is willing to admit.

Cinematic 13-second raw handheld streetwear fashion video shot on 8mm vintage fisheye lens, ultra wide 16mm perspective, fast-paced energetic Tokyo urban night sequence, dynamic raw Y2K hip hop aesthetic, 16:9 aspect ratio. 0 to 2 seconds: Extreme close-up opening through a narrow vertical zipper slit in pitch blackness, revealing a young Asian man wearing a black streetwear bucket hat and embroidered jacket crouching down. The camera rapidly pans out through the slit into a super wide fisheye low-angle perspective looking up at two young men standing in a glowing Tokyo street corner at night surrounded by tall neon skyscrapers under dark skies. 2 to 4 seconds: Quick handheld whip-pan to the right down a neon lit sidewalk. A young black man with braided hair wearing a bright royal blue sweater with yellow lettering and a green paisley scarf steps casually through the frame while gesturing at the camera. Fast rotational motion blur transition sweeps across the screen. 4 to 6 seconds: Scene cuts smoothly inside a bright Japanese claw machine arcade filled with pastel pink interior walls and fluorescent overhead lights. The camera tracks backward as the young man in the blue sweater walks forward holding a small pink object to his ear while smiling and talking. An animated stylized liquid frame graphic swirls briefly around the edges of the frame. 6 to 8 seconds: Seamless cut to a retro mint green fast-food diner booth with warm overhead hanging lamp lighting. Low-angle wide view on the white tabletop as the man in the blue sweater reaches his hand directly toward the camera glass offering food. A quick dynamic psychedelic thermal infrared visual effect flashes over the screen for half a second showing vivid heat map colors of red, neon green, and deep blue before snapping back to real life. 8 to 10 seconds: The second man wearing the black jacket suddenly pops up from the diner table with an energetic reaction, holding a drink cup as the camera sweeps low. Fast cut to outdoors on a night street where both men walk along the pavement and high five each other in motion under glowing urban shop signs. 10 to 13 seconds: Camera drops to an extreme ground-level fisheye lens pointing straight up from the asphalt of a narrow alleyway. Both men loom over the lens under vertical glowing Japanese neon signboards. One subject stoops down and reaches his palm directly over the camera lens, causing a rapid spinning rotational motion blur that wipes the screen to black. Visual details: Authentic analog film grain, natural subtle lens distortion, realistic handheld camera shake, rich contrasty neon lighting, crisp reflections on pavement, high energy streetwear styling, realistic 90s hip hop music video camera direction. Audio includes an upbeat boom bap 90s hip hop beat with heavy bass, classic vinyl scratches, energetic vocal chops, synth accents, dynamic street ambiance, arcade sound effects, and clean rhythmic percussion layered throughout.

4 levers

Make it yours

What is safe to change. Most libraries publish only this list, which is why so many copied prompts come back worse than the original.

  1. 01

    The lens

    8mm vintage fisheye, ultra-wide 16mm perspective. Named at the top and again on the asphalt. If you swap it for a clean 35mm the six locations stop being one film and become a travel ad.

  2. 02

    The bookends

    Opens through a zipper, ends with a palm on the lens. Both are wipes made of objects, not of dissolves. Change the ending to a logo and you have thrown away the only joke.

  3. 03

    The second man

    Blue sweater, yellow lettering, green paisley scarf. He is the moving subject for four of the six blocks. Recast him and rewrite every garment or the arcade and the diner invent a stranger.

  4. 04

    The thermal flash

    Half a second of infrared at 6–8s. It is a spice beat. Delete it if you want a cleaner clip; lengthen it and the diner never comes back.

Three ways to break it

  • Smoothing the camera

    The prompt asks for raw handheld shake and analog grain. Stabilise the language and H3 will give you a modern brand film, which is the opposite of a Y2K tape.

  • Dropping the zipper or the palm

    Those two objects are the edit. Without them the six locations are just six locations, and thirteen seconds is not enough locations to get away with that.

  • Writing real shop names into the neon

    The prompt wants Japanese signboards, not readable brands. Named stores are a rights problem and a texture problem — the model will misspell them.

What it does

What the tokyo fisheye streetwear template does

Most montage prompts list locations and hope the model invents a way from one to the next. This one names the transitions as things in the world: a zipper slit, a whip-pan, a liquid frame, a thermal flash, a palm on the glass. The locations are what those objects reveal.

The 0–2 / 2–4 / 4–6 / 6–8 / 8–10 / 10–13 grid is two-second blocks plus a three-second landing. Two seconds is enough for one camera idea and not enough for a scene. That is why the arcade is a walk toward camera and not a game being played.

Two men, two palettes. Black bucket hat and embroidered jacket against royal blue, yellow type, green paisley. The prompt never lets them wear the same colours, so a fisheye that bends the whole street still has two readable figures.

The audio paragraph is unusually complete: boom-bap, vinyl scratches, vocal chops, arcade, street. H3 will generate a bed whether you ask or not; writing this one is how you stop it inventing a lo-fi piano over a hip-hop picture.

Thirteen seconds is inside H3's range, and the prompt states it. Set the duration chip to 13s so the text and the request agree. Sixteen-by-nine is stated twice. Do not run it vertical — the ground-level ending is a sky full of signs, and 9:16 crops the men off the sides.

The source is a reference render from another model. The fisheye will come back softer on H3 than on the clip beside this prompt. What transfers is the zipper, the palm, and the two-second grid, not the exact grain.

Everything here runs in text to video.

6 questions

Tokyo fisheye streetwear — common questions

  • 01

    Why 13 seconds and not 15?

    Because the source is timed to 13. Padding to 15 adds a seventh location the prompt did not write, and the palm wipe arrives late.

  • 02

    Can I change the city?

    Yes, if you keep neon, a cramped alley, and signboards that are not readable brands. The fisheye needs vertical lights to bend.

  • 03

    Do I need both men?

    The high-five and the ending need two bodies. A one-man version has to rewrite 8–13s, not just delete a description.

  • 04

    Will the thermal flash survive?

    Sometimes. It is a half-second effect in a two-second block. If it disappears, the diner still works; do not add a second effect to compensate.

  • 05

    Should I mention a brand of camera?

    No. "8mm vintage fisheye" is the look. A brand name is a logo the model will try to paint on the lens.

  • 06

    What does one run cost?

    Thirteen seconds at 2K is $1.69 at the published per-second rate.

Copy it, change three things, run it.

Every character is on this page. 13s at 2K costs $1.69.

Generate free