Utopai-X Flash vs Flash Lite vs the market

A hands-on test of the public Video API at video.internal-utopai.com, run on 12 Nov 2026 from a script (POST /v1/videos, polling until done). Every Utopai clip here is the raw API output at 768p (1344×768), 24 fps, 16:9, with native audio, no retakes and no cherry-picking: one generation per prompt and model.

Speed and cost, measured

ModelClipRunsRangeMedianPrice per clip
Flash10 s · 768p827–48 s31 sUS$ 1.31
Flash Lite10 s · 768p816–29 s23 sUS$ 0.63
Flash6 s · 768p1819–26 s23 sUS$ 0.79
Flash Lite6 s · 768p1812–26 s15 sUS$ 0.38

Generation time is the API's started_at → finished_at; jobs were sent three at a time (the workspace concurrency), so there was no queue. List price: Flash US$ 0.131/s, Flash Lite US$ 0.063/s at 768p. 52 jobs, 52 succeeded, US$ 36.58 spent.

Output file, as delivered

H.264 1344×768 at 24 fps (1.75:1, not exactly 16:9), 0.3–2.6 Mbps; AAC stereo at 32 kHz; a 10 s request returns 10.125 s.

New prompts (N1–N8) · 10 s · Flash vs Flash Lite

N1 · Physics: fire, oil, rice in the air; hands 10 s

Utopai X Flash48 s to generate · US$ 1.31
Utopai X Flash Lite29 s to generate · US$ 0.63
Prompt sent to Utopai (1239 chars)
ShotID: uxt-n1-wok-fuego-v1

A night street-food stall in Bangkok. Handheld documentary camera at chest height, 35mm lens, one meter from the cook, the framing lagging slightly behind his hands. The cook: fifties, sweat on his temples, grey T-shirt, tossing rice and shrimp in a black carbon-steel wok over a roaring gas burner.

0–3 s: he ladles oil into the wok; it shimmers and starts to smoke.
3–6 s: he tips the wok toward the burner and a column of fire flares half a meter high, lighting his face orange for one second, then collapses.
6–10 s: he flips the rice twice with a flick of the wrist, every grain arcs and lands back in the wok, then he slides it onto a white plate held in his other hand and taps the ladle twice on the wok's rim.

Physics: rice and shrimp follow real ballistic arcs and land; the flame rises from the burner and the wok's lip, never from his hands; steam and smoke drift up and out of frame. Shallow depth of field, T1.8: background traffic and neon stay soft bokeh.

Audio: only diegetic sound — the roar of the burner, the hiss and crackle of oil, the scrape of the steel ladle, two sharp taps on the rim, distant scooters. No music. Single continuous take, no cuts. No music, no soundtrack, no score.

N2 · Dialogue with lip sync: who speaks, living eyes, rain 10 s

Utopai X Flash31 s to generate · US$ 1.31
Utopai X Flash Lite16 s to generate · US$ 0.63
Prompt sent to Utopai (1138 chars)
ShotID: uxt-n2-dialogo-auto-v1

Night, inside a parked car in heavy rain. Static camera on the dashboard, 28mm lens, facing the two front seats, both faces in frame. Driver's seat, frame left: a woman, thirties, wet hair, denim jacket, both hands on the steering wheel. Passenger seat, frame right: a man, forties, grey stubble, a folded paper envelope on his knee. Orange streetlight through the rain-streaked side windows throws moving rain shadows across their faces.

0–4 s: she looks straight ahead and says quietly: "You said it would be one night."
4–8 s: he turns the envelope over twice in his fingers, without looking at her, and answers: "It was. I just didn't say which one."
8–10 s: she exhales through her nose, almost a laugh, and turns her head to look at him; he finally meets her eyes.

Lip movement matches every syllable. Each one speaks only their own line; the listener's mouth stays closed. Eyes alive: blinks, small breaths, micro-movements.

Audio: rain drumming on the roof, faint engine idle, the envelope's paper rustling, their voices close and dry. No music. One continuous take, no cuts. No music, no score.

N3 · Three shots with hard cuts: wardrobe and screen-direction continuity 10 s

Utopai X Flash31 s to generate · US$ 1.31
Utopai X Flash Lite24 s to generate · US$ 0.63
Prompt sent to Utopai (1277 chars)
ShotID: uxt-n3-ascensor-cortes-v1

A three-shot sequence with hard cuts: same woman, same lobby, consistent screen direction — she always moves from frame left to frame right. The woman: late twenties, short black bob, red wool coat, brown leather tote bag on her right shoulder. A marble office lobby at 8 a.m., cold daylight through tall glass.

Shot 1 (0–3 s), wide, camera low at the far end of the lobby: she bursts through the revolving door and runs toward the elevators on the right, heels clicking, the tote bouncing on her hip.
Shot 2 (3–6 s), medium tracking shot alongside her at waist height, 50mm: she runs past a security desk, a guard half-rises from his chair; she raises one hand. Ahead, the steel elevator doors are already sliding closed.
Shot 3 (6–10 s), close-up from inside the elevator, 85mm, T1.4: her hand slips into the last ten centimeters of the gap, the doors bump her fingers and slide back open; she steps in breathing hard, a strand of hair stuck to her lip, and gives a breathless half-smile to someone off-screen left.

Continuity: same red coat, same tote on the right shoulder, same hair in all three shots.

Audio: heels on marble, the revolving-door whoosh, the elevator chime, the door motor, her breathing. No music. No music, no score.

N4 · Legible, correctly spelled text in frame 10 s

Utopai X Flash35 s to generate · US$ 1.31
Utopai X Flash Lite26 s to generate · US$ 0.63
Prompt sent to Utopai (1061 chars)
ShotID: uxt-n4-texto-cartel-v1

A small family bakery at dawn, seen from the sidewalk. Slow push-in on a 40mm lens, eye level, from three meters to one meter. A baker in a white apron, sixties, flour on his forearms, hangs a hand-painted wooden sign on a hook inside the shop window. The sign reads, in large dark-green brush letters: "FRESH BREAD 6 AM". Leaning against the glass below it, a slate board in white chalk: "Sourdough $4". Behind the glass: golden loaves on wooden shelves, warm tungsten light inside; outside, cool blue dawn, and the glass reflects the empty street.

0–3 s: wide on the storefront as he hangs the sign.
3–7 s: the camera pushes in; he straightens the sign with two fingers, then wipes his hands on the apron.
7–10 s: the push-in ends on the sign, fully legible and sharp, the baker's silhouette soft behind it.

Text is spelled exactly as written above, with no other letters or words anywhere in frame.

Audio: the hook squeaking, wood knocking on glass, a distant early bus, one bird. No music. Single take. No music, no score.

N5 · Crowd without clones or merging; scale; aerial move 10 s

Utopai X Flash30 s to generate · US$ 1.31
Utopai X Flash Lite22 s to generate · US$ 0.63
Prompt sent to Utopai (1038 chars)
ShotID: uxt-n5-multitud-mercado-v1

Aerial drone shot, 60 meters high, camera tilted 45° down, slow forward glide with a gentle orbit to the right, over a crowded open-air fruit market in a Mexican town plaza at 7 a.m. Hundreds of people, each on their own path and at their own pace: vendors unloading crates from pickup trucks, shoppers carrying bags, a boy on a bicycle weaving through, a woman pulling a cart. Rows of orange, yellow and pink tarp canopies; long morning shadows falling to the left. Scale anchor: each tarp stall is three meters wide; people are small but readable.

Over the 10 seconds the drone crosses the plaza and reveals a white colonial church on the far side, its bell tower catching the first sunlight.

No two people move in sync; no repeated faces or outfits; every person stays one person — no merging, no popping in or out.

Audio: the market's murmur rising from below, a pickup engine, a vendor calling out prices, church bells starting at 8 s, light wind. No music. Continuous take. No music, no score.

N6 · Animal motion: four-beat gallop, spray, visible speed 10 s

Utopai X Flash28 s to generate · US$ 1.31
Utopai X Flash Lite25 s to generate · US$ 0.63
Prompt sent to Utopai (1005 chars)
ShotID: uxt-n6-caballo-agua-v1

A dark bay horse and rider gallop through shallow sea water along a beach at sunset. Camera on a vehicle running parallel, ten meters to the horse's left, 85mm lens, matching speed, slight vibration. The rider: a woman in her forties, white linen shirt, leaning forward in the saddle.

The horse moves in a correct four-beat gallop: legs in the right sequence, hooves striking the water and throwing fans of spray that catch the backlight, muscles rippling, mane and tail streaming back. Speed is visible: waves and wet sand stream from right to left behind them, spray hangs and falls.

0–6 s: steady parallel tracking at full gallop.
6–10 s: the horse slows to a canter, then a trot; the spray dies down; the rider sits back and pats the horse's neck once.

Strong backlight, the sun low on the horizon behind them, sea haze.

Audio: hooves splashing, the horse's heavy breathing, saddle leather creaking, surf, wind. No music. Single continuous take. No music, no score.

N7 · Extreme close-up: skin, a tear, subtle acting, T1.4 focus 10 s

Utopai X Flash34 s to generate · US$ 1.31
Utopai X Flash Lite22 s to generate · US$ 0.63
Prompt sent to Utopai (949 chars)
ShotID: uxt-n7-primerisimo-llanto-v1

Extreme close-up of a man's face, seventies, deep wrinkles, white stubble, pale blue eyes, filling the frame from eyebrows to chin. 100mm macro lens, T1.4, focus on the eyes; the ears are already soft. Warm window light from frame left; the other side falls into shadow. He is reading a letter held just below frame; we never see it.

0–3 s: his eyes move left to right along the lines, steady.
3–6 s: the eyes stop on one line, go back and reread it; his jaw tightens, his lower lip presses up.
6–10 s: tears pool along the lower eyelid and a single tear runs down beside his nose; he blinks twice, slowly, breathes out through his mouth, and one corner of his mouth lifts in a small smile.

Skin with real pores and fine hairs, the wet shine of the eye. Tiny handheld breathing in the frame.

Audio: room tone, a clock ticking, paper rustling once, his breath catching. No music. One take. No music, no score.

N8 · One-take gimbal shot: coherent space, exposure change, wine in glasses 10 s

Utopai X Flash27 s to generate · US$ 1.31
Utopai X Flash Lite20 s to generate · US$ 0.63
Prompt sent to Utopai (1006 chars)
ShotID: uxt-n8-plano-secuencia-v1

One continuous gimbal shot, 24mm lens, chest height, following a young waiter from behind as he carries a tray with three glasses of red wine. The camera stays half a meter behind his left shoulder the whole time.

0–3 s: a dim, warm restaurant kitchen, steam, cooks plating; he pushes through a swinging door with his shoulder.
3–6 s: the candlelit dining room; he weaves between tables.
6–10 s: he steps out through open glass doors onto a sunny terrace over the sea at noon; the exposure adapts over one second from warm dark to bright daylight; he sets the three glasses down on a table where an elderly couple waits.

The wine stays in the glasses and sloshes realistically; the tray stays level. The space is physically coherent: kitchen → door → dining room → terrace in one line, no cuts.

Audio: kitchen clatter and sizzle, the swinging door, quiet dinner chatter and cutlery, then waves and gulls outside, glasses set down on wood. No music. No music, no score.

Raw Kitchen · 6 s · Utopai vs Seedance 2.5, MiniMax H3 and Wan 3.0

rk-01-pelican-v2 6 s

Utopai X Flash25 s to generate · US$ 0.79
Utopai X Flash Lite19 s to generate · US$ 0.38
Seedance 2.5 (PAI Pro)1080p · 6 s
MiniMax H3 (PAI)1080p · 6 s

Not identical: Utopai accepts up to 2,000 characters, so the opening of the look paragraph, shared by every shot, was condensed (the same way in all 13 affected shots). Audio, animal description and shot-by-shot timeline are unchanged.

Prompt sent to Utopai (1997 chars)
ShotID: rk-01-pelican-v2

AUDIO
No music of any kind. Only diegetic sound: a water pour, pouch splash, wingbeats, webbed feet on marble, a pelican grunt and room tone. The influencer reacts with her face and body only.

LOOK, CAMERA AND PERFORMANCE
Photographic live-action cooking vlog in a bright luxury kitchen: pale-stone walls, a long white marble island, brass taps, tall back windows. Strong backlight rims fur and feathers and makes flour, steam and droplets glow in the air. Lens at T1.4–T1.8: razor-thin focus on the animal, soft bright bokeh behind. Handheld camera, never still: breathing sway, imperfect framing, late pans, held low and close. Blockbuster scale: the key moment ramps into slow motion; every impact is physical. The pelican keeps true anatomy, weight, white-grey plumage and behavior; its own bill and throat pouch do the pouring. The adult influencer, in soft focus behind the island, reacts big and believable—eyes wide, a delighted grin—and treats the animal as normal cooking company.

TIMELINE — 6 SECONDS, 4 HARD CUTS
0:00–0:01.3 — Low wide shot from counter height: the pelican lands on the island edge, huge wings spread against the backlit window, a puff of flour lifting off the marble in its wingbeat. SFX: heavy wingbeats, webbed feet slapping stone.
0:01.3–0:02.8 — Macro profile: the pelican tilts its enormous bill over a clear glass measuring cup, the pouch sagging with water, a first silver stream breaking from the bill tip.
0:02.8–0:04.6 — Slow motion, detail through the glass: water climbs the cup, bubbles rising, the surface stopping exactly on the printed 250 ml line, one last drop hanging in the backlight before it falls.
0:04.6–0:06.0 — Low counter-level angle: the pelican lifts its head, snaps its bill shut once and settles its feathers; behind it in soft focus the influencer nods with a proud grin. SFX: bill clack, a low pelican grunt.

FINAL AUDIO
No music of any kind. The last cut ends on the gag's own sound and kitchen ambience.
Original prompt used in PAI (Seedance 2.5 and the variant) (2232 chars)
ShotID: rk-01-pelican-v2

AUDIO
No music of any kind. Only diegetic sound: a water pour, pouch splash, wingbeats, webbed feet on marble, a pelican grunt and kitchen room tone. The influencer reacts with her face and body only.

LOOK, CAMERA AND PERFORMANCE
Photographic live-action cooking vlog in a bright, oversized luxury kitchen: pale-stone walls, a long white marble island, brushed brass taps, tall windows along the back wall. Strong backlight pours through those windows, rimming fur and feathers and turning flour, steam and droplets into glowing particles in the air. Fast lens wide open at T1.4–T1.8: razor-thin focus on the animal, backgrounds melt into soft bright bokeh. Human handheld camera, never still: tremors, breathing sway, imperfect framing, delayed pans and small focus corrections, held low and close to the animal. Blockbuster scale inside a kitchen: the key moment ramps into slow motion, every impact is physical and visible. The real pelican keeps true anatomy, size, weight, white-grey plumage and species behavior; its actual bill and throat pouch do the pouring. The adult influencer, in soft focus behind the island, reacts big and believable—eyes wide, a delighted grin—and treats the animal as normal cooking company.

TIMELINE — 6 SECONDS, 4 HARD CUTS
0:00–0:01.3 — Low wide shot from counter height: the pelican lands on the island edge, huge wings spread against the backlit window, a puff of flour lifting off the marble in its wingbeat. SFX: heavy wingbeats, webbed feet slapping stone.
0:01.3–0:02.8 — Macro profile: the pelican tilts its enormous bill over a clear glass measuring cup, the pouch sagging with water, a first silver stream breaking from the bill tip.
0:02.8–0:04.6 — Slow motion, detail through the glass: water climbs the cup, bubbles rising, the surface stopping exactly on the printed 250 ml line, one last drop hanging in the backlight before it falls.
0:04.6–0:06.0 — Low counter-level angle: the pelican lifts its head, snaps its bill shut once and settles its feathers; behind it in soft focus the influencer nods with a proud grin. SFX: bill clack, a low pelican grunt.

FINAL AUDIO
No music of any kind. The last cut ends on the gag's own sound and kitchen ambience.

rk-02-cow-v2 6 s

Utopai X Flash25 s to generate · US$ 0.79
Utopai X Flash Lite16 s to generate · US$ 0.38
Seedance 2.5 (PAI Pro)1080p · 6 s
Wan 3.0 (PAI)1080p · 6 s

Not identical: Utopai accepts up to 2,000 characters, so the opening of the look paragraph, shared by every shot, was condensed (the same way in all 13 affected shots). Audio, animal description and shot-by-shot timeline are unchanged.

Prompt sent to Utopai (1972 chars)
ShotID: rk-02-cow-v2

AUDIO
No music of any kind. Only diegetic sound: hooves on stone, a deep moo, a jet of milk hitting glass, froth rising, a cow's breath and room tone. The influencer reacts with her face and body only.

LOOK, CAMERA AND PERFORMANCE
Photographic live-action cooking vlog in a bright luxury kitchen: pale-stone walls, a long white marble island, brass taps, tall back windows. Strong backlight rims fur and feathers and makes flour, steam and droplets glow in the air. Lens at T1.4–T1.8: razor-thin focus on the animal, soft bright bokeh behind. Handheld camera, never still: breathing sway, imperfect framing, late pans, held low and close. Blockbuster scale: the key moment ramps into slow motion; every impact is physical. The cow keeps true anatomy, weight, black-and-white hide and behavior; milk falls from its own udder by gravity into a glass jug on a low stool beneath it. Scale anchor: the cow's back rises above the marble island; the jug reaches its knee. The adult influencer, in soft focus behind the island, reacts big and believable—eyes wide, a delighted grin—and treats the animal as normal cooking company.

TIMELINE — 6 SECONDS, 4 HARD CUTS
0:00–0:01.3 — Low wide shot: the cow stands beside the island, its flank filling the frame against the backlit window; it turns its head toward the lens and moos.
0:01.3–0:02.8 — Macro on the udder: a teat tightens and a first thick white jet shoots straight down into the jug, splashing the glass.
0:02.8–0:04.6 — Slow motion detail through the glass: the jet hits the pool, froth climbs the wall, droplets hang in the backlight, and the milk rises to the brim of the jug.
0:04.6–0:06.0 — Medium shot: the cow chews and flicks an ear, the last drops falling; the influencer reaches into frame, lifts the full jug and grins wide at the cow. SFX: the stream easing, glass ringing, a soft snort.

FINAL AUDIO
No music of any kind. The last cut ends on the gag's own sound and kitchen ambience.
Original prompt used in PAI (Seedance 2.5 and the variant) (2321 chars)
ShotID: rk-02-cow-v2

AUDIO
No music of any kind. Only diegetic sound: hooves on stone, a deep moo, a jet of milk hitting glass, froth rising, a cow's breath and kitchen room tone. The influencer reacts with her face and body only.

LOOK, CAMERA AND PERFORMANCE
Photographic live-action cooking vlog in a bright, oversized luxury kitchen: pale-stone walls, a long white marble island, brushed brass taps, tall windows along the back wall. Strong backlight pours through those windows, rimming fur and feathers and turning flour, steam and droplets into glowing particles in the air. Fast lens wide open at T1.4–T1.8: razor-thin focus on the animal, backgrounds melt into soft bright bokeh. Human handheld camera, never still: tremors, breathing sway, imperfect framing, delayed pans and small focus corrections, held low and close to the animal. Blockbuster scale inside a kitchen: the key moment ramps into slow motion, every impact is physical and visible. The real cow keeps true anatomy, size, weight, black-and-white hide and species behavior; its actual udder gives the milk, which falls by gravity into a glass jug standing on a low stool directly beneath it. Scale anchor: the cow's back rises above the marble island; the jug reaches its knee. The adult influencer, in soft focus behind the island, reacts big and believable—eyes wide, a delighted grin—and treats the animal as normal cooking company.

TIMELINE — 6 SECONDS, 4 HARD CUTS
0:00–0:01.3 — Low wide shot: the cow stands beside the island, its flank filling the frame against the backlit window, a clear glass jug on a low stool under its udder; it turns its head toward the lens and moos. SFX: hooves on stone, a deep moo.
0:01.3–0:02.8 — Macro on the udder: a teat tightens and a first thick white jet shoots straight down into the jug, splashing the glass.
0:02.8–0:04.6 — Slow motion detail through the glass: the jet hits the pool, froth climbs the wall, droplets hang in the backlight, and the milk rises to the brim of the jug.
0:04.6–0:06.0 — Medium shot: the cow chews and flicks an ear, the last drops falling; the influencer reaches into frame, lifts the full jug and grins wide at the cow. SFX: the stream easing, glass ringing, a soft snort.

FINAL AUDIO
No music of any kind. The last cut ends on the gag's own sound and kitchen ambience.

rk-03-hen-v1 6 s

Utopai X Flash25 s to generate · US$ 0.79
Utopai X Flash Lite14 s to generate · US$ 0.38
Seedance 2.5 (PAI Pro)1080p · 6 s
Wan 3.0 (PAI)1080p · 6 s

Not identical: Utopai accepts up to 2,000 characters, so the opening of the look paragraph, shared by every shot, was condensed (the same way in all 13 affected shots). Audio, animal description and shot-by-shot timeline are unchanged.

Prompt sent to Utopai (1879 chars)
ShotID: rk-03-hen-v1

AUDIO
No music of any kind. Only diegetic sound: clucking, claws gripping the bowl rim, feathers rustling, the egg landing in flour with a soft thud and kitchen room tone. The influencer reacts with her face and body only.

LOOK, CAMERA AND PERFORMANCE
Photographic live-action cooking vlog in a bright luxury kitchen: pale-stone walls, a long white marble island, brass taps, tall back windows. Strong backlight rims fur and feathers and makes flour, steam and droplets glow in the air. Lens at T1.4–T1.8: razor-thin focus on the animal, soft bright bokeh behind. Handheld camera, never still: breathing sway, imperfect framing, late pans, held low and close. Blockbuster scale: the key moment ramps into slow motion; every impact is physical. The real hen keeps true anatomy, size, weight, russet plumage and species behavior; its actual body lays the egg. The adult influencer, in soft focus behind the island, reacts big and believable—eyes wide, a delighted grin—and treats the animal as normal cooking company.

TIMELINE — 6 SECONDS, 4 HARD CUTS
0:00–0:01.3 — Low macro: the hen's scaly feet grip the rim of a wide ceramic mixing bowl, flour dust glowing in the backlight behind it. SFX: claws scraping ceramic, a soft cluck.
0:01.3–0:02.8 — Medium shot at bowl level: the hen crouches, feathers fluffing out, eyes half closing, the bowl rocking a few millimetres under her weight.
0:02.8–0:04.6 — Slow motion detail: a fresh brown egg drops from under her, falls through the shaft of backlight and lands in the flour with a soft puff that rises like a cloud.
0:04.6–0:06.0 — Low wide shot: the hen hops off the rim with one flap and struts away along the marble; behind the bowl the influencer gasps and grins. SFX: a proud cluck, wings, feet on stone.

FINAL AUDIO
No music of any kind. The last cut ends on the gag's own sound and kitchen ambience.
Original prompt used in PAI (Seedance 2.5 and the variant) (2084 chars)
ShotID: rk-03-hen-v1

AUDIO
No music of any kind. Only diegetic sound: clucking, claws gripping the bowl rim, feathers rustling, the egg landing in flour with a soft thud and kitchen room tone. The influencer reacts with her face and body only.

LOOK, CAMERA AND PERFORMANCE
Photographic live-action cooking vlog in a bright, oversized luxury kitchen: pale-stone walls, a long white marble island, brushed brass taps, tall windows along the back wall. Strong backlight pours through those windows, rimming fur and feathers and turning flour, steam and droplets into glowing particles in the air. Fast lens wide open at T1.4–T1.8: razor-thin focus on the animal, backgrounds melt into soft bright bokeh. Human handheld camera, never still: tremors, breathing sway, imperfect framing, delayed pans and small focus corrections, held low and close to the animal. Blockbuster scale inside a kitchen: the key moment ramps into slow motion, every impact is physical and visible. The real hen keeps true anatomy, size, weight, russet plumage and species behavior; its actual body lays the egg. The adult influencer, in soft focus behind the island, reacts big and believable—eyes wide, a delighted grin—and treats the animal as normal cooking company.

TIMELINE — 6 SECONDS, 4 HARD CUTS
0:00–0:01.3 — Low macro: the hen's scaly feet grip the rim of a wide ceramic mixing bowl, flour dust glowing in the backlight behind it. SFX: claws scraping ceramic, a soft cluck.
0:01.3–0:02.8 — Medium shot at bowl level: the hen crouches, feathers fluffing out, eyes half closing, the bowl rocking a few millimetres under her weight.
0:02.8–0:04.6 — Slow motion detail: a fresh brown egg drops from under her, falls through the shaft of backlight and lands in the flour with a soft puff that rises like a cloud.
0:04.6–0:06.0 — Low wide shot: the hen hops off the rim with one flap and struts away along the marble; behind the bowl the influencer gasps and grins. SFX: a proud cluck, wings, feet on stone.

FINAL AUDIO
No music of any kind. The last cut ends on the gag's own sound and kitchen ambience.

rk-04-octopus-v1 6 s

Utopai X Flash23 s to generate · US$ 0.79
Utopai X Flash Lite17 s to generate · US$ 0.38
Seedance 2.5 (PAI Pro)1080p · 6 s
MiniMax H3 (PAI)1080p · 6 s

Not identical: Utopai accepts up to 2,000 characters, so the opening of the look paragraph, shared by every shot, was condensed (the same way in all 13 affected shots). Audio, animal description and shot-by-shot timeline are unchanged.

Prompt sent to Utopai (1799 chars)
ShotID: rk-04-octopus-v1

AUDIO
No music of any kind. Only diegetic sound: wet suckers on marble, squelching dough, slapping tentacles, the bowl knocking on stone and kitchen room tone. The influencer reacts with her face and body only.

LOOK, CAMERA AND PERFORMANCE
Photographic live-action cooking vlog in a bright luxury kitchen: pale-stone walls, a long white marble island, brass taps, tall back windows. Strong backlight rims fur and feathers and makes flour, steam and droplets glow in the air. Lens at T1.4–T1.8: razor-thin focus on the animal, soft bright bokeh behind. Handheld camera, never still: breathing sway, imperfect framing, late pans, held low and close. Blockbuster scale: the key moment ramps into slow motion; every impact is physical. The real octopus keeps true anatomy, size, weight, mottled wet skin and species behavior; its actual tentacles and suckers do the mixing. The influencer stays beyond the frame; the animal owns the shot.

TIMELINE — 6 SECONDS, 4 HARD CUTS
0:00–0:01.3 — Low wide shot: the octopus hauls itself up over the edge of the ceramic bowl, suckers popping onto the marble, its skin rippling in the backlight. SFX: wet suckers, a slap on stone.
0:01.3–0:02.8 — Macro: three tentacles plunge into the flour, water and egg, curling and folding, the mixture climbing the bowl wall.
0:02.8–0:04.6 — Slow motion detail: dough strands stretch between two tentacles, flour bursts upward through the light, a sucker releases with a visible pop.
0:04.6–0:06.0 — Low angle at counter level: the bowl rocks and knocks on the marble, the island shudders, and the octopus's eye sits in razor focus as a fourth tentacle slides in to join. SFX: squelch, bowl knocking.

FINAL AUDIO
No music of any kind. The last cut ends on the gag's own sound and kitchen ambience.
Original prompt used in PAI (Seedance 2.5 and the variant) (2004 chars)
ShotID: rk-04-octopus-v1

AUDIO
No music of any kind. Only diegetic sound: wet suckers on marble, squelching dough, slapping tentacles, the bowl knocking on stone and kitchen room tone. The influencer reacts with her face and body only.

LOOK, CAMERA AND PERFORMANCE
Photographic live-action cooking vlog in a bright, oversized luxury kitchen: pale-stone walls, a long white marble island, brushed brass taps, tall windows along the back wall. Strong backlight pours through those windows, rimming fur and feathers and turning flour, steam and droplets into glowing particles in the air. Fast lens wide open at T1.4–T1.8: razor-thin focus on the animal, backgrounds melt into soft bright bokeh. Human handheld camera, never still: tremors, breathing sway, imperfect framing, delayed pans and small focus corrections, held low and close to the animal. Blockbuster scale inside a kitchen: the key moment ramps into slow motion, every impact is physical and visible. The real octopus keeps true anatomy, size, weight, mottled wet skin and species behavior; its actual tentacles and suckers do the mixing. The influencer stays beyond the frame; the animal owns the shot.

TIMELINE — 6 SECONDS, 4 HARD CUTS
0:00–0:01.3 — Low wide shot: the octopus hauls itself up over the edge of the ceramic bowl, suckers popping onto the marble, its skin rippling in the backlight. SFX: wet suckers, a slap on stone.
0:01.3–0:02.8 — Macro: three tentacles plunge into the flour, water and egg, curling and folding, the mixture climbing the bowl wall.
0:02.8–0:04.6 — Slow motion detail: dough strands stretch between two tentacles, flour bursts upward through the light, a sucker releases with a visible pop.
0:04.6–0:06.0 — Low angle at counter level: the bowl rocks and knocks on the marble, the island shudders, and the octopus's eye sits in razor focus as a fourth tentacle slides in to join. SFX: squelch, bowl knocking.

FINAL AUDIO
No music of any kind. The last cut ends on the gag's own sound and kitchen ambience.

rk-05-elephant-v1 6 s

Utopai X Flash22 s to generate · US$ 0.79
Utopai X Flash Lite12 s to generate · US$ 0.38
Seedance 2.5 (PAI Pro)1080p · 6 s
Wan 3.0 (PAI)1080p · 6 s

Not identical: Utopai accepts up to 2,000 characters, so the opening of the look paragraph, shared by every shot, was condensed (the same way in all 13 affected shots). Audio, animal description and shot-by-shot timeline are unchanged.

Prompt sent to Utopai (1989 chars)
ShotID: rk-05-elephant-v1

AUDIO
No music of any kind. Only diegetic sound: the trunk breathing, dough thudding and folding, the counter creaking, flour hissing and kitchen room tone. The influencer reacts with her face and body only.

LOOK, CAMERA AND PERFORMANCE
Photographic live-action cooking vlog in a bright luxury kitchen: pale-stone walls, a long white marble island, brass taps, tall back windows. Strong backlight rims fur and feathers and makes flour, steam and droplets glow in the air. Lens at T1.4–T1.8: razor-thin focus on the animal, soft bright bokeh behind. Handheld camera, never still: breathing sway, imperfect framing, late pans, held low and close. Blockbuster scale: the key moment ramps into slow motion; every impact is physical. The real elephant keeps true anatomy, size, weight, grey wrinkled skin and species behavior; its actual trunk does the kneading. Scale anchor: the elephant's head fills the open back window; its trunk reaches across the island to the dough. The adult influencer, in soft focus behind the island, reacts big and believable—eyes wide, a delighted grin—and treats the animal as normal cooking company.

TIMELINE — 6 SECONDS, 4 HARD CUTS
0:00–0:01.3 — Wide shot from the island: the elephant's head fills the open back window, backlit, and its trunk reaches in across the marble toward the ball of dough. SFX: a deep breath through the trunk.
0:01.3–0:02.8 — Detail: the trunk tip presses into the dough, which bulges and gives, a puff of flour rising around it.
0:02.8–0:04.6 — Slow motion macro: the trunk folds the dough over itself and pushes down, the counter shuddering, a cloud of flour drifting through the backlight.
0:04.6–0:06.0 — Low angle: the trunk lifts away leaving a smooth folded dough; the influencer stands small in the frame looking up at the window with a huge grin. SFX: the trunk withdrawing, the counter settling.

FINAL AUDIO
No music of any kind. The last cut ends on the gag's own sound and kitchen ambience.
Original prompt used in PAI (Seedance 2.5 and the variant) (2194 chars)
ShotID: rk-05-elephant-v1

AUDIO
No music of any kind. Only diegetic sound: the trunk breathing, dough thudding and folding, the counter creaking, flour hissing and kitchen room tone. The influencer reacts with her face and body only.

LOOK, CAMERA AND PERFORMANCE
Photographic live-action cooking vlog in a bright, oversized luxury kitchen: pale-stone walls, a long white marble island, brushed brass taps, tall windows along the back wall. Strong backlight pours through those windows, rimming fur and feathers and turning flour, steam and droplets into glowing particles in the air. Fast lens wide open at T1.4–T1.8: razor-thin focus on the animal, backgrounds melt into soft bright bokeh. Human handheld camera, never still: tremors, breathing sway, imperfect framing, delayed pans and small focus corrections, held low and close to the animal. Blockbuster scale inside a kitchen: the key moment ramps into slow motion, every impact is physical and visible. The real elephant keeps true anatomy, size, weight, grey wrinkled skin and species behavior; its actual trunk does the kneading. Scale anchor: the elephant's head fills the open back window; its trunk reaches across the island to the dough. The adult influencer, in soft focus behind the island, reacts big and believable—eyes wide, a delighted grin—and treats the animal as normal cooking company.

TIMELINE — 6 SECONDS, 4 HARD CUTS
0:00–0:01.3 — Wide shot from the island: the elephant's head fills the open back window, backlit, and its trunk reaches in across the marble toward the ball of dough. SFX: a deep breath through the trunk.
0:01.3–0:02.8 — Detail: the trunk tip presses into the dough, which bulges and gives, a puff of flour rising around it.
0:02.8–0:04.6 — Slow motion macro: the trunk folds the dough over itself and pushes down, the counter shuddering, a cloud of flour drifting through the backlight.
0:04.6–0:06.0 — Low angle: the trunk lifts away leaving a smooth folded dough; the influencer stands small in the frame looking up at the window with a huge grin. SFX: the trunk withdrawing, the counter settling.

FINAL AUDIO
No music of any kind. The last cut ends on the gag's own sound and kitchen ambience.

rk-06-armadillo-v1 6 s

Utopai X Flash24 s to generate · US$ 0.79
Utopai X Flash Lite12 s to generate · US$ 0.38
Seedance 2.5 (PAI Pro)1080p · 6 s
MiniMax H3 (PAI)1080p · 6 s
Identical prompt, word for word, on all four engines (1975 chars)
ShotID: rk-06-armadillo-v1

AUDIO
No music of any kind. Only diegetic sound: shell plates clicking, dough squeaking flat, flour hissing, claws on marble and kitchen room tone. The influencer reacts with her face and body only.

LOOK, CAMERA AND PERFORMANCE
Photographic live-action cooking vlog in a bright, oversized luxury kitchen: pale-stone walls, a long white marble island, brushed brass taps, tall windows along the back wall. Strong backlight pours through those windows, rimming fur and feathers and turning flour, steam and droplets into glowing particles in the air. Fast lens wide open at T1.4–T1.8: razor-thin focus on the animal, backgrounds melt into soft bright bokeh. Human handheld camera, never still: tremors, breathing sway, imperfect framing, delayed pans and small focus corrections, held low and close to the animal. Blockbuster scale inside a kitchen: the key moment ramps into slow motion, every impact is physical and visible. The real armadillo keeps true anatomy, size, weight, banded armored shell and species behavior; its actual curled shell does the rolling. The influencer stays beyond the frame; the animal owns the shot.

TIMELINE — 6 SECONDS, 4 HARD CUTS
0:00–0:01.3 — Macro: the armadillo stands at the near edge of a flat sheet of dough, nose twitching, flour on its snout glowing in the backlight. SFX: snuffling, claws on stone.
0:01.3–0:02.8 — Medium shot at counter level: it tucks its head and tail and curls into a tight ball, armor plates clicking shut.
0:02.8–0:04.6 — Slow motion low angle: the ball rolls across the dough, the dough flattening and spreading under it, flour spraying up on both sides through the light, the counter vibrating.
0:04.6–0:06.0 — Detail from the far edge: the ball slows, uncurls, and the armadillo blinks over a perfectly flat round of dough stretched out behind it. SFX: plates unclicking, a snort.

FINAL AUDIO
No music of any kind. The last cut ends on the gag's own sound and kitchen ambience.

rk-07-ostrich-v1 6 s

Utopai X Flash19 s to generate · US$ 0.79
Utopai X Flash Lite13 s to generate · US$ 0.38
Seedance 2.5 (PAI Pro)1080p · 6 s
MiniMax H3 (PAI)1080p · 6 s

Not identical: Utopai accepts up to 2,000 characters, so the opening of the look paragraph, shared by every shot, was condensed (the same way in all 13 affected shots). Audio, animal description and shot-by-shot timeline are unchanged.

Prompt sent to Utopai (1834 chars)
ShotID: rk-07-ostrich-v1

AUDIO
No music of any kind. Only diegetic sound: rapid beak strikes on dough, flour puffing, the counter drumming, a low ostrich boom and kitchen room tone. The influencer reacts with her face and body only.

LOOK, CAMERA AND PERFORMANCE
Photographic live-action cooking vlog in a bright luxury kitchen: pale-stone walls, a long white marble island, brass taps, tall back windows. Strong backlight rims fur and feathers and makes flour, steam and droplets glow in the air. Lens at T1.4–T1.8: razor-thin focus on the animal, soft bright bokeh behind. Handheld camera, never still: breathing sway, imperfect framing, late pans, held low and close. Blockbuster scale: the key moment ramps into slow motion; every impact is physical. The real ostrich keeps true anatomy, size, weight, black and white plumage and species behavior; its actual beak does the perforating. The adult influencer, in soft focus behind the island, reacts big and believable—eyes wide, a delighted grin—and treats the animal as normal cooking company.

TIMELINE — 6 SECONDS, 4 HARD CUTS
0:00–0:01.3 — Low wide shot: the ostrich's long neck drops from above into frame, its head stopping a hand's width over the flat dough, eyelashes backlit. SFX: a low boom from its throat.
0:01.3–0:02.8 — Macro: the beak strikes the dough and lifts, strikes and lifts, each hit leaving a clean hole.
0:02.8–0:04.6 — Slow motion detail: a machine-gun row of pecks marches across the dough, a flour puff rising from each strike, the counter drumming under the blows.
0:04.6–0:06.0 — Medium shot: the head rises, the ostrich blinks once over an evenly perforated dough; behind it the influencer flinches, then laughs. SFX: the last peck, feathers settling.

FINAL AUDIO
No music of any kind. The last cut ends on the gag's own sound and kitchen ambience.
Original prompt used in PAI (Seedance 2.5 and the variant) (2039 chars)
ShotID: rk-07-ostrich-v1

AUDIO
No music of any kind. Only diegetic sound: rapid beak strikes on dough, flour puffing, the counter drumming, a low ostrich boom and kitchen room tone. The influencer reacts with her face and body only.

LOOK, CAMERA AND PERFORMANCE
Photographic live-action cooking vlog in a bright, oversized luxury kitchen: pale-stone walls, a long white marble island, brushed brass taps, tall windows along the back wall. Strong backlight pours through those windows, rimming fur and feathers and turning flour, steam and droplets into glowing particles in the air. Fast lens wide open at T1.4–T1.8: razor-thin focus on the animal, backgrounds melt into soft bright bokeh. Human handheld camera, never still: tremors, breathing sway, imperfect framing, delayed pans and small focus corrections, held low and close to the animal. Blockbuster scale inside a kitchen: the key moment ramps into slow motion, every impact is physical and visible. The real ostrich keeps true anatomy, size, weight, black and white plumage and species behavior; its actual beak does the perforating. The adult influencer, in soft focus behind the island, reacts big and believable—eyes wide, a delighted grin—and treats the animal as normal cooking company.

TIMELINE — 6 SECONDS, 4 HARD CUTS
0:00–0:01.3 — Low wide shot: the ostrich's long neck drops from above into frame, its head stopping a hand's width over the flat dough, eyelashes backlit. SFX: a low boom from its throat.
0:01.3–0:02.8 — Macro: the beak strikes the dough and lifts, strikes and lifts, each hit leaving a clean hole.
0:02.8–0:04.6 — Slow motion detail: a machine-gun row of pecks marches across the dough, a flour puff rising from each strike, the counter drumming under the blows.
0:04.6–0:06.0 — Medium shot: the head rises, the ostrich blinks once over an evenly perforated dough; behind it the influencer flinches, then laughs. SFX: the last peck, feathers settling.

FINAL AUDIO
No music of any kind. The last cut ends on the gag's own sound and kitchen ambience.

rk-08-monkey-v1 6 s

Utopai X Flash21 s to generate · US$ 0.79
Utopai X Flash Lite13 s to generate · US$ 0.38
Seedance 2.5 (PAI Pro)1080p · 6 s
MiniMax H3 (PAI)1080p · 6 s

Not identical: Utopai accepts up to 2,000 characters, so the opening of the look paragraph, shared by every shot, was condensed (the same way in all 13 affected shots). Audio, animal description and shot-by-shot timeline are unchanged.

Prompt sent to Utopai (1801 chars)
ShotID: rk-08-monkey-v1

AUDIO
No music of any kind. Only diegetic sound: fingers squeaking on glass, the lid seal popping, the lid ringing on marble, a monkey chatter and kitchen room tone. The influencer reacts with her face and body only.

LOOK, CAMERA AND PERFORMANCE
Photographic live-action cooking vlog in a bright luxury kitchen: pale-stone walls, a long white marble island, brass taps, tall back windows. Strong backlight rims fur and feathers and makes flour, steam and droplets glow in the air. Lens at T1.4–T1.8: razor-thin focus on the animal, soft bright bokeh behind. Handheld camera, never still: breathing sway, imperfect framing, late pans, held low and close. Blockbuster scale: the key moment ramps into slow motion; every impact is physical. The real monkey keeps true anatomy, size, weight, brown fur and species behavior; its actual hands and grip do the opening. The adult influencer, in soft focus behind the island, reacts big and believable—eyes wide, a delighted grin—and treats the animal as normal cooking company.

TIMELINE — 6 SECONDS, 4 HARD CUTS
0:00–0:01.3 — Detail: the monkey's hands close around a big glass jar of tomatoes, fur rimmed in the backlight, fingers squeaking on the glass. SFX: squeak, a quick chatter.
0:01.3–0:02.8 — Medium low shot: it braces the jar against its chest and twists; the seal pops with a clear snap.
0:02.8–0:04.6 — Slow motion macro: the metal lid spins off into the air, a single bead of tomato juice flying beside it, both catching the light.
0:04.6–0:06.0 — Wide shot: the monkey holds the open jar out toward the counter and smacks its lips; the influencer claps once with a wide grin. SFX: the lid landing and ringing on marble.

FINAL AUDIO
No music of any kind. The last cut ends on the gag's own sound and kitchen ambience.
Original prompt used in PAI (Seedance 2.5 and the variant) (2006 chars)
ShotID: rk-08-monkey-v1

AUDIO
No music of any kind. Only diegetic sound: fingers squeaking on glass, the lid seal popping, the lid ringing on marble, a monkey chatter and kitchen room tone. The influencer reacts with her face and body only.

LOOK, CAMERA AND PERFORMANCE
Photographic live-action cooking vlog in a bright, oversized luxury kitchen: pale-stone walls, a long white marble island, brushed brass taps, tall windows along the back wall. Strong backlight pours through those windows, rimming fur and feathers and turning flour, steam and droplets into glowing particles in the air. Fast lens wide open at T1.4–T1.8: razor-thin focus on the animal, backgrounds melt into soft bright bokeh. Human handheld camera, never still: tremors, breathing sway, imperfect framing, delayed pans and small focus corrections, held low and close to the animal. Blockbuster scale inside a kitchen: the key moment ramps into slow motion, every impact is physical and visible. The real monkey keeps true anatomy, size, weight, brown fur and species behavior; its actual hands and grip do the opening. The adult influencer, in soft focus behind the island, reacts big and believable—eyes wide, a delighted grin—and treats the animal as normal cooking company.

TIMELINE — 6 SECONDS, 4 HARD CUTS
0:00–0:01.3 — Detail: the monkey's hands close around a big glass jar of tomatoes, fur rimmed in the backlight, fingers squeaking on the glass. SFX: squeak, a quick chatter.
0:01.3–0:02.8 — Medium low shot: it braces the jar against its chest and twists; the seal pops with a clear snap.
0:02.8–0:04.6 — Slow motion macro: the metal lid spins off into the air, a single bead of tomato juice flying beside it, both catching the light.
0:04.6–0:06.0 — Wide shot: the monkey holds the open jar out toward the counter and smacks its lips; the influencer claps once with a wide grin. SFX: the lid landing and ringing on marble.

FINAL AUDIO
No music of any kind. The last cut ends on the gag's own sound and kitchen ambience.

rk-09-flamingo-v1 6 s

Utopai X Flash20 s to generate · US$ 0.79
Utopai X Flash Lite14 s to generate · US$ 0.38
Seedance 2.5 (PAI Pro)1080p · 6 s
MiniMax H3 (PAI)1080p · 6 s
Identical prompt, word for word, on all four engines (1952 chars)
ShotID: rk-09-flamingo-v1

AUDIO
No music of any kind. Only diegetic sound: a whirring beak in sauce, sauce sloshing against glass, a wet spatter on the lens, a flamingo honk and kitchen room tone. The influencer reacts with her face and body only.

LOOK, CAMERA AND PERFORMANCE
Photographic live-action cooking vlog in a bright, oversized luxury kitchen: pale-stone walls, a long white marble island, brushed brass taps, tall windows along the back wall. Strong backlight pours through those windows, rimming fur and feathers and turning flour, steam and droplets into glowing particles in the air. Fast lens wide open at T1.4–T1.8: razor-thin focus on the animal, backgrounds melt into soft bright bokeh. Human handheld camera, never still: tremors, breathing sway, imperfect framing, delayed pans and small focus corrections, held low and close to the animal. Blockbuster scale inside a kitchen: the key moment ramps into slow motion, every impact is physical and visible. The real flamingo keeps true anatomy, size, weight, pink plumage and species behavior; its actual curved beak and rapid head motion do the blending. The influencer stays beyond the frame; the animal owns the shot.

TIMELINE — 6 SECONDS, 4 HARD CUTS
0:00–0:01.3 — Low wide shot: the flamingo steps up to the island on one leg, its neck curving down into a tall glass jar of peeled tomatoes, pink feathers glowing in the backlight. SFX: a honk, feet on stone.
0:01.3–0:02.8 — Macro: the beak plunges under the surface and the head whirs side to side at high speed, the sauce churning.
0:02.8–0:04.6 — Slow motion detail: a red vortex climbs the inside of the glass, a fine spatter hitting the lens, a drop hanging in the light.
0:04.6–0:06.0 — Detail: the flamingo lifts its head, smooth sauce dripping from the beak tip, and shakes its neck once. SFX: the drip, a satisfied honk.

FINAL AUDIO
No music of any kind. The last cut ends on the gag's own sound and kitchen ambience.

rk-10-woodpecker-v1 6 s

Utopai X Flash23 s to generate · US$ 0.79
Utopai X Flash Lite15 s to generate · US$ 0.38
Seedance 2.5 (PAI Pro)1080p · 6 s
MiniMax H3 (PAI)1080p · 6 s
Identical prompt, word for word, on all four engines (1980 chars)
ShotID: rk-10-woodpecker-v1

AUDIO
No music of any kind. Only diegetic sound: staccato beak hammering on wood, garlic skins crackling, mince scattering, a sharp chirp and kitchen room tone. The influencer reacts with her face and body only.

LOOK, CAMERA AND PERFORMANCE
Photographic live-action cooking vlog in a bright, oversized luxury kitchen: pale-stone walls, a long white marble island, brushed brass taps, tall windows along the back wall. Strong backlight pours through those windows, rimming fur and feathers and turning flour, steam and droplets into glowing particles in the air. Fast lens wide open at T1.4–T1.8: razor-thin focus on the animal, backgrounds melt into soft bright bokeh. Human handheld camera, never still: tremors, breathing sway, imperfect framing, delayed pans and small focus corrections, held low and close to the animal. Blockbuster scale inside a kitchen: the key moment ramps into slow motion, every impact is physical and visible. The real woodpecker keeps true anatomy, size, weight, red crest and species behavior; its actual beak does the mincing. The adult influencer, in soft focus behind the island, reacts big and believable—eyes wide, a delighted grin—and treats the animal as normal cooking company.

TIMELINE — 6 SECONDS, 4 HARD CUTS
0:00–0:01.3 — Macro: the woodpecker lands on the edge of a wooden board beside a garlic bulb, red crest blazing in the backlight. SFX: wings, claws on wood.
0:01.3–0:02.8 — Detail: the beak hammers into the garlic at absurd speed, papery skins crackling and flying.
0:02.8–0:04.6 — Slow motion macro: cloves split into fine mince, pieces tumbling through the light, the board vibrating under each strike.
0:04.6–0:06.0 — Low angle: a neat pile of minced garlic; the woodpecker shakes its head and a flake lands on the influencer's nose as she grins in soft focus. SFX: a final rapid drumroll, a chirp.

FINAL AUDIO
No music of any kind. The last cut ends on the gag's own sound and kitchen ambience.

rk-11-bear-v1 6 s

Utopai X Flash23 s to generate · US$ 0.79
Utopai X Flash Lite26 s to generate · US$ 0.38
Seedance 2.5 (PAI Pro)1080p · 6 s
MiniMax H3 (PAI)1080p · 6 s

Not identical: Utopai accepts up to 2,000 characters, so the opening of the look paragraph, shared by every shot, was condensed (the same way in all 13 affected shots). Audio, animal description and shot-by-shot timeline are unchanged.

Prompt sent to Utopai (1897 chars)
ShotID: rk-11-bear-v1

AUDIO
No music of any kind. Only diegetic sound: a heavy paw thudding on marble, tomatoes bursting, walnut shells cracking, bowls jumping, a low huff and kitchen room tone. The influencer reacts with her face and body only.

LOOK, CAMERA AND PERFORMANCE
Photographic live-action cooking vlog in a bright luxury kitchen: pale-stone walls, a long white marble island, brass taps, tall back windows. Strong backlight rims fur and feathers and makes flour, steam and droplets glow in the air. Lens at T1.4–T1.8: razor-thin focus on the animal, soft bright bokeh behind. Handheld camera, never still: breathing sway, imperfect framing, late pans, held low and close. Blockbuster scale: the key moment ramps into slow motion; every impact is physical. The real bear keeps true anatomy, size, weight, thick brown fur and species behavior; its actual paw and claws do the crushing. Scale anchor: standing behind the island, the bear's shoulders rise above the influencer's head. The adult influencer, in soft focus behind the island, reacts big and believable—eyes wide, a delighted grin—and treats the animal as normal cooking company.

TIMELINE — 6 SECONDS, 4 HARD CUTS
0:00–0:01.3 — Low wide shot: the bear rises behind the island, one massive paw lifted against the backlit window, fur glowing at the edges. SFX: a low huff, claws clicking.
0:01.3–0:02.8 — Medium shot: the paw comes down on a heap of tomatoes and walnuts; the counter shudders and nearby bowls jump.
0:02.8–0:04.6 — Slow motion macro: tomato skins burst, red juice sheets outward, walnut shells split and shards tumble through the light.
0:04.6–0:06.0 — Medium: the bear lifts the dripping paw and licks it once; the influencer, small in soft focus, stares wide-eyed then grins. SFX: wet drip, a satisfied grunt.

FINAL AUDIO
No music of any kind. The last cut ends on the gag's own sound and kitchen ambience.
Original prompt used in PAI (Seedance 2.5 and the variant) (2102 chars)
ShotID: rk-11-bear-v1

AUDIO
No music of any kind. Only diegetic sound: a heavy paw thudding on marble, tomatoes bursting, walnut shells cracking, bowls jumping, a low huff and kitchen room tone. The influencer reacts with her face and body only.

LOOK, CAMERA AND PERFORMANCE
Photographic live-action cooking vlog in a bright, oversized luxury kitchen: pale-stone walls, a long white marble island, brushed brass taps, tall windows along the back wall. Strong backlight pours through those windows, rimming fur and feathers and turning flour, steam and droplets into glowing particles in the air. Fast lens wide open at T1.4–T1.8: razor-thin focus on the animal, backgrounds melt into soft bright bokeh. Human handheld camera, never still: tremors, breathing sway, imperfect framing, delayed pans and small focus corrections, held low and close to the animal. Blockbuster scale inside a kitchen: the key moment ramps into slow motion, every impact is physical and visible. The real bear keeps true anatomy, size, weight, thick brown fur and species behavior; its actual paw and claws do the crushing. Scale anchor: standing behind the island, the bear's shoulders rise above the influencer's head. The adult influencer, in soft focus behind the island, reacts big and believable—eyes wide, a delighted grin—and treats the animal as normal cooking company.

TIMELINE — 6 SECONDS, 4 HARD CUTS
0:00–0:01.3 — Low wide shot: the bear rises behind the island, one massive paw lifted against the backlit window, fur glowing at the edges. SFX: a low huff, claws clicking.
0:01.3–0:02.8 — Medium shot: the paw comes down on a heap of tomatoes and walnuts; the counter shudders and nearby bowls jump.
0:02.8–0:04.6 — Slow motion macro: tomato skins burst, red juice sheets outward, walnut shells split and shards tumble through the light.
0:04.6–0:06.0 — Medium: the bear lifts the dripping paw and licks it once; the influencer, small in soft focus, stares wide-eyed then grins. SFX: wet drip, a satisfied grunt.

FINAL AUDIO
No music of any kind. The last cut ends on the gag's own sound and kitchen ambience.

rk-12-parrot-v1 6 s

Utopai X Flash26 s to generate · US$ 0.79
Utopai X Flash Lite17 s to generate · US$ 0.38
Seedance 2.5 (PAI Pro)1080p · 6 s
Wan 3.0 (PAI)1080p · 6 s

Not identical: Utopai accepts up to 2,000 characters, so the opening of the look paragraph, shared by every shot, was condensed (the same way in all 13 affected shots). Audio, animal description and shot-by-shot timeline are unchanged.

Prompt sent to Utopai (1822 chars)
ShotID: rk-12-parrot-v1

AUDIO
No music of any kind. Only diegetic sound: walnut shells cracking, shell halves ticking on marble, nut meat landing in a mortar, a parrot shriek and kitchen room tone. The influencer reacts with her face and body only.

LOOK, CAMERA AND PERFORMANCE
Photographic live-action cooking vlog in a bright luxury kitchen: pale-stone walls, a long white marble island, brass taps, tall back windows. Strong backlight rims fur and feathers and makes flour, steam and droplets glow in the air. Lens at T1.4–T1.8: razor-thin focus on the animal, soft bright bokeh behind. Handheld camera, never still: breathing sway, imperfect framing, late pans, held low and close. Blockbuster scale: the key moment ramps into slow motion; every impact is physical. The real parrot keeps true anatomy, size, weight, scarlet and blue plumage and species behavior; its actual hooked beak and foot do the cracking. The adult influencer, in soft focus behind the island, reacts big and believable—eyes wide, a delighted grin—and treats the animal as normal cooking company.

TIMELINE — 6 SECONDS, 4 HARD CUTS
0:00–0:01.3 — Macro: the parrot's foot pins a walnut on the marble beside a stone mortar of basil, its hooked beak closing over the shell, feathers blazing in the backlight. SFX: claws, a soft crack.
0:01.3–0:02.8 — Slow motion detail: the shell splits cleanly, both halves falling away and ticking onto the counter.
0:02.8–0:04.6 — Low angle: the parrot drops the nut meat into the mortar and snatches a second walnut, cracking it in one bite.
0:04.6–0:06.0 — Wide shot: the parrot shrieks once and fluffs its feathers over the filled mortar; the influencer laughs behind it in soft focus. SFX: shriek, feathers.

FINAL AUDIO
No music of any kind. The last cut ends on the gag's own sound and kitchen ambience.
Original prompt used in PAI (Seedance 2.5 and the variant) (2027 chars)
ShotID: rk-12-parrot-v1

AUDIO
No music of any kind. Only diegetic sound: walnut shells cracking, shell halves ticking on marble, nut meat landing in a mortar, a parrot shriek and kitchen room tone. The influencer reacts with her face and body only.

LOOK, CAMERA AND PERFORMANCE
Photographic live-action cooking vlog in a bright, oversized luxury kitchen: pale-stone walls, a long white marble island, brushed brass taps, tall windows along the back wall. Strong backlight pours through those windows, rimming fur and feathers and turning flour, steam and droplets into glowing particles in the air. Fast lens wide open at T1.4–T1.8: razor-thin focus on the animal, backgrounds melt into soft bright bokeh. Human handheld camera, never still: tremors, breathing sway, imperfect framing, delayed pans and small focus corrections, held low and close to the animal. Blockbuster scale inside a kitchen: the key moment ramps into slow motion, every impact is physical and visible. The real parrot keeps true anatomy, size, weight, scarlet and blue plumage and species behavior; its actual hooked beak and foot do the cracking. The adult influencer, in soft focus behind the island, reacts big and believable—eyes wide, a delighted grin—and treats the animal as normal cooking company.

TIMELINE — 6 SECONDS, 4 HARD CUTS
0:00–0:01.3 — Macro: the parrot's foot pins a walnut on the marble beside a stone mortar of basil, its hooked beak closing over the shell, feathers blazing in the backlight. SFX: claws, a soft crack.
0:01.3–0:02.8 — Slow motion detail: the shell splits cleanly, both halves falling away and ticking onto the counter.
0:02.8–0:04.6 — Low angle: the parrot drops the nut meat into the mortar and snatches a second walnut, cracking it in one bite.
0:04.6–0:06.0 — Wide shot: the parrot shrieks once and fluffs its feathers over the filled mortar; the influencer laughs behind it in soft focus. SFX: shriek, feathers.

FINAL AUDIO
No music of any kind. The last cut ends on the gag's own sound and kitchen ambience.

rk-13-beaver-v1 6 s

Utopai X Flash26 s to generate · US$ 0.79
Utopai X Flash Lite16 s to generate · US$ 0.38
Seedance 2.5 (PAI Pro)1080p · 6 s
MiniMax H3 (PAI)1080p · 6 s
Identical prompt, word for word, on all four engines (1989 chars)
ShotID: rk-13-beaver-v1

AUDIO
No music of any kind. Only diegetic sound: incisors scraping cheese, shavings showering onto dough, a flat tail slapping marble, a grunt and kitchen room tone. The influencer reacts with her face and body only.

LOOK, CAMERA AND PERFORMANCE
Photographic live-action cooking vlog in a bright, oversized luxury kitchen: pale-stone walls, a long white marble island, brushed brass taps, tall windows along the back wall. Strong backlight pours through those windows, rimming fur and feathers and turning flour, steam and droplets into glowing particles in the air. Fast lens wide open at T1.4–T1.8: razor-thin focus on the animal, backgrounds melt into soft bright bokeh. Human handheld camera, never still: tremors, breathing sway, imperfect framing, delayed pans and small focus corrections, held low and close to the animal. Blockbuster scale inside a kitchen: the key moment ramps into slow motion, every impact is physical and visible. The real beaver keeps true anatomy, size, weight, wet brown fur and species behavior; its actual incisors and flat tail do the work. The adult influencer, in soft focus behind the island, reacts big and believable—eyes wide, a delighted grin—and treats the animal as normal cooking company.

TIMELINE — 6 SECONDS, 4 HARD CUTS
0:00–0:01.3 — Low wide shot: the beaver clambers onto the island beside a cheese block the size of its own body, whiskers backlit, water dripping from its fur. SFX: claws on stone, a grunt.
0:01.3–0:02.8 — Macro: its orange incisors bite and scrape along the block, cheese curling off in ribbons.
0:02.8–0:04.6 — Slow motion detail: a shower of shavings cascades through the light onto the sauced dough, drifting like snow.
0:04.6–0:06.0 — Medium: the beaver slaps its flat tail on the marble with a crack; the influencer flinches and grins behind the pizza. SFX: tail slap, cheese still falling.

FINAL AUDIO
No music of any kind. The last cut ends on the gag's own sound and kitchen ambience.

rk-14-crab-v1 6 s

Utopai X Flash23 s to generate · US$ 0.79
Utopai X Flash Lite17 s to generate · US$ 0.38
Seedance 2.5 (PAI Pro)1080p · 6 s
Wan 3.0 (PAI)1080p · 6 s

Not identical: Utopai accepts up to 2,000 characters, so the opening of the look paragraph, shared by every shot, was condensed (the same way in all 13 affected shots). Audio, animal description and shot-by-shot timeline are unchanged.

Prompt sent to Utopai (1860 chars)
ShotID: rk-14-crab-v1

AUDIO
No music of any kind. Only diegetic sound: claws snipping stems, leaves landing on sauce, claws cutting through crust, crumbs crackling, claw clicks and kitchen room tone. The influencer reacts with her face and body only.

LOOK, CAMERA AND PERFORMANCE
Photographic live-action cooking vlog in a bright luxury kitchen: pale-stone walls, a long white marble island, brass taps, tall back windows. Strong backlight rims fur and feathers and makes flour, steam and droplets glow in the air. Lens at T1.4–T1.8: razor-thin focus on the animal, soft bright bokeh behind. Handheld camera, never still: breathing sway, imperfect framing, late pans, held low and close. Blockbuster scale: the key moment ramps into slow motion; every impact is physical. The real crab keeps true anatomy, size, weight, orange-red shell and species behavior; its actual claws do the snipping and the slicing. The shot jumps in time from raw to baked pizza. The influencer stays beyond the frame; the animal owns the shot.

TIMELINE — 6 SECONDS, 5 HARD CUTS
0:00–0:01.2 — Macro: the crab's claws scissor through fresh basil stems, leaves dropping onto the red sauce of the raw pizza, shell gleaming in the backlight. SFX: snip, leaves landing.
0:01.2–0:02.4 — Slow motion detail: one leaf drifts down through the light and lands flat on the sauce.
0:02.4–0:03.6 — Low wide shot, time jump: the baked pizza steams on a board, crust blistered; the crab scuttles sideways onto it.
0:03.6–0:05.0 — Slow motion macro: the claws close through the crust in a straight radial cut, crumbs crackling and lifting through the steam.
0:05.0–0:06.0 — Detail from above: eight perfect slices; the crab raises both claws and clicks them twice. SFX: crust cracking, claw clicks.

FINAL AUDIO
No music of any kind. The last cut ends on the gag's own sound and kitchen ambience.
Original prompt used in PAI (Seedance 2.5 and the variant) (2065 chars)
ShotID: rk-14-crab-v1

AUDIO
No music of any kind. Only diegetic sound: claws snipping stems, leaves landing on sauce, claws cutting through crust, crumbs crackling, claw clicks and kitchen room tone. The influencer reacts with her face and body only.

LOOK, CAMERA AND PERFORMANCE
Photographic live-action cooking vlog in a bright, oversized luxury kitchen: pale-stone walls, a long white marble island, brushed brass taps, tall windows along the back wall. Strong backlight pours through those windows, rimming fur and feathers and turning flour, steam and droplets into glowing particles in the air. Fast lens wide open at T1.4–T1.8: razor-thin focus on the animal, backgrounds melt into soft bright bokeh. Human handheld camera, never still: tremors, breathing sway, imperfect framing, delayed pans and small focus corrections, held low and close to the animal. Blockbuster scale inside a kitchen: the key moment ramps into slow motion, every impact is physical and visible. The real crab keeps true anatomy, size, weight, orange-red shell and species behavior; its actual claws do the snipping and the slicing. The shot jumps in time from raw to baked pizza. The influencer stays beyond the frame; the animal owns the shot.

TIMELINE — 6 SECONDS, 5 HARD CUTS
0:00–0:01.2 — Macro: the crab's claws scissor through fresh basil stems, leaves dropping onto the red sauce of the raw pizza, shell gleaming in the backlight. SFX: snip, leaves landing.
0:01.2–0:02.4 — Slow motion detail: one leaf drifts down through the light and lands flat on the sauce.
0:02.4–0:03.6 — Low wide shot, time jump: the baked pizza steams on a board, crust blistered; the crab scuttles sideways onto it.
0:03.6–0:05.0 — Slow motion macro: the claws close through the crust in a straight radial cut, crumbs crackling and lifting through the steam.
0:05.0–0:06.0 — Detail from above: eight perfect slices; the crab raises both claws and clicks them twice. SFX: crust cracking, claw clicks.

FINAL AUDIO
No music of any kind. The last cut ends on the gag's own sound and kitchen ambience.

rk-15-chameleon-v1 6 s

Utopai X Flash24 s to generate · US$ 0.79
Utopai X Flash Lite23 s to generate · US$ 0.38
Seedance 2.5 (PAI Pro)1080p · 6 s
Wan 3.0 (PAI)1080p · 6 s

Not identical: Utopai accepts up to 2,000 characters, so the opening of the look paragraph, shared by every shot, was condensed (the same way in all 13 affected shots). Audio, animal description and shot-by-shot timeline are unchanged.

Prompt sent to Utopai (1942 chars)
ShotID: rk-15-chameleon-v1

AUDIO
No music of any kind. Only diegetic sound: a tongue snapping out, the jar knocked off wood, glass landing on marble, the tongue retracting with a wet snap and kitchen room tone. The influencer reacts with her face and body only.

LOOK, CAMERA AND PERFORMANCE
Photographic live-action cooking vlog in a bright luxury kitchen: pale-stone walls, a long white marble island, brass taps, tall back windows. Strong backlight rims fur and feathers and makes flour, steam and droplets glow in the air. Lens at T1.4–T1.8: razor-thin focus on the animal, soft bright bokeh behind. Handheld camera, never still: breathing sway, imperfect framing, late pans, held low and close. Blockbuster scale: the key moment ramps into slow motion; every impact is physical. The real chameleon keeps true anatomy, size, weight, green scaled skin and species behavior; its actual tongue does the retrieving. Geography: the high open shelf of spice jars is above the island, lit from the window beside it. The adult influencer, in soft focus behind the island, reacts big and believable—eyes wide, a delighted grin—and treats the animal as normal cooking company.

TIMELINE — 6 SECONDS, 4 HARD CUTS
0:00–0:01.3 — Low wide shot: the chameleon sits on the marble, both eyes swiveling up toward a high open shelf of glass seasoning jars glowing in the backlight. SFX: a soft click of claws.
0:01.3–0:02.8 — Slow motion macro: the tongue fires upward out of frame, a long wet ribbon stretching through the light.
0:02.8–0:04.6 — Detail on the shelf: the tongue tip hits a jar of chili flakes and yanks it off the wood, the jar sailing down.
0:04.6–0:06.0 — Medium shot: the jar lands upright on the marble as the tongue snaps back into the mouth; the eyes swivel; behind, the influencer claps with a huge grin. SFX: glass landing, wet snap.

FINAL AUDIO
No music of any kind. The last cut ends on the gag's own sound and kitchen ambience.
Original prompt used in PAI (Seedance 2.5 and the variant) (2147 chars)
ShotID: rk-15-chameleon-v1

AUDIO
No music of any kind. Only diegetic sound: a tongue snapping out, the jar knocked off wood, glass landing on marble, the tongue retracting with a wet snap and kitchen room tone. The influencer reacts with her face and body only.

LOOK, CAMERA AND PERFORMANCE
Photographic live-action cooking vlog in a bright, oversized luxury kitchen: pale-stone walls, a long white marble island, brushed brass taps, tall windows along the back wall. Strong backlight pours through those windows, rimming fur and feathers and turning flour, steam and droplets into glowing particles in the air. Fast lens wide open at T1.4–T1.8: razor-thin focus on the animal, backgrounds melt into soft bright bokeh. Human handheld camera, never still: tremors, breathing sway, imperfect framing, delayed pans and small focus corrections, held low and close to the animal. Blockbuster scale inside a kitchen: the key moment ramps into slow motion, every impact is physical and visible. The real chameleon keeps true anatomy, size, weight, green scaled skin and species behavior; its actual tongue does the retrieving. Geography: the high open shelf of spice jars is above the island, lit from the window beside it. The adult influencer, in soft focus behind the island, reacts big and believable—eyes wide, a delighted grin—and treats the animal as normal cooking company.

TIMELINE — 6 SECONDS, 4 HARD CUTS
0:00–0:01.3 — Low wide shot: the chameleon sits on the marble, both eyes swiveling up toward a high open shelf of glass seasoning jars glowing in the backlight. SFX: a soft click of claws.
0:01.3–0:02.8 — Slow motion macro: the tongue fires upward out of frame, a long wet ribbon stretching through the light.
0:02.8–0:04.6 — Detail on the shelf: the tongue tip hits a jar of chili flakes and yanks it off the wood, the jar sailing down.
0:04.6–0:06.0 — Medium shot: the jar lands upright on the marble as the tongue snaps back into the mouth; the eyes swivel; behind, the influencer claps with a huge grin. SFX: glass landing, wet snap.

FINAL AUDIO
No music of any kind. The last cut ends on the gag's own sound and kitchen ambience.

rk-16-snake-v1 6 s

Utopai X Flash20 s to generate · US$ 0.79
Utopai X Flash Lite14 s to generate · US$ 0.38
Seedance 2.5 (PAI Pro)1080p · 6 s
MiniMax H3 (PAI)1080p · 6 s
Identical prompt, word for word, on all four engines (1915 chars)
ShotID: rk-16-snake-v1

AUDIO
No music of any kind. Only diegetic sound: scales sliding on fabric, the bag creaking under pressure, ricotta squeezing out in a wet ribbon, a tongue flick and kitchen room tone. The influencer reacts with her face and body only.

LOOK, CAMERA AND PERFORMANCE
Photographic live-action cooking vlog in a bright, oversized luxury kitchen: pale-stone walls, a long white marble island, brushed brass taps, tall windows along the back wall. Strong backlight pours through those windows, rimming fur and feathers and turning flour, steam and droplets into glowing particles in the air. Fast lens wide open at T1.4–T1.8: razor-thin focus on the animal, backgrounds melt into soft bright bokeh. Human handheld camera, never still: tremors, breathing sway, imperfect framing, delayed pans and small focus corrections, held low and close to the animal. Blockbuster scale inside a kitchen: the key moment ramps into slow motion, every impact is physical and visible. The real snake keeps true anatomy, size, weight, patterned scales and species behavior; its actual coils do the squeezing. The influencer stays beyond the frame; the animal owns the shot.

TIMELINE — 6 SECONDS, 4 HARD CUTS
0:00–0:01.3 — Macro: the snake's coils are wrapped around a bulging cloth piping bag on the marble, scales sliding over the fabric in the backlight. SFX: scales on cloth.
0:01.3–0:02.8 — Medium low shot: the coils tighten, the bag creaks and the nozzle lifts toward the open edge of the pizza crust.
0:02.8–0:04.6 — Slow motion detail: a thick white ribbon of ricotta flows from the nozzle into the crust's open rim, the dough swelling as it fills.
0:04.6–0:06.0 — Detail: the snake's head follows the rim to the end of the fill and its tongue flicks once over the seam. SFX: the last squeeze, tongue flick.

FINAL AUDIO
No music of any kind. The last cut ends on the gag's own sound and kitchen ambience.

rk-18-duck-v1 6 s

Utopai X Flash20 s to generate · US$ 0.79
Utopai X Flash Lite15 s to generate · US$ 0.38
Seedance 2.5 (PAI Pro)1080p · 6 s
MiniMax H3 (PAI)1080p · 6 s

Not identical: Utopai accepts up to 2,000 characters, so the opening of the look paragraph, shared by every shot, was condensed (the same way in all 13 affected shots). Audio, animal description and shot-by-shot timeline are unchanged.

Prompt sent to Utopai (1897 chars)
ShotID: rk-18-duck-v1

AUDIO
No music of any kind. Only diegetic sound: a fine spray of water, droplets ticking on dough, a violent wing shake, water sheeting over marble and glass, a startled quack and kitchen room tone. The influencer reacts with her face and body only.

LOOK, CAMERA AND PERFORMANCE
Photographic live-action cooking vlog in a bright luxury kitchen: pale-stone walls, a long white marble island, brass taps, tall back windows. Strong backlight rims fur and feathers and makes flour, steam and droplets glow in the air. Lens at T1.4–T1.8: razor-thin focus on the animal, soft bright bokeh behind. Handheld camera, never still: breathing sway, imperfect framing, late pans, held low and close. Blockbuster scale: the key moment ramps into slow motion; every impact is physical. The real duck keeps true anatomy, size, weight, wet white plumage and species behavior; its actual bill and wings do the spraying and the drenching. The adult influencer, in soft focus behind the island, reacts big and believable—eyes wide, a delighted grin—and treats the animal as normal cooking company.

TIMELINE — 6 SECONDS, 4 HARD CUTS
0:00–0:01.3 — Macro: the duck's bill dribbles a fine, delicate mist over the stretched dough, each droplet glowing in the backlight. SFX: a soft spray, ticking drops.
0:01.3–0:02.8 — Slow motion detail: droplets hang in the light and settle on the flour-dusted surface.
0:02.8–0:04.6 — Medium shot: the duck shakes its head and then its whole body and wings, water sheeting off in every direction, bowls and glasses hit by the spray.
0:04.6–0:06.0 — Wide shot: the entire kitchen is drenched, water running down the window, droplets on the lens; the camera jolts; the influencer, soaked, freezes wide-eyed, then bursts out laughing. SFX: quack, water everywhere.

FINAL AUDIO
No music of any kind. The last cut ends on the gag's own sound and kitchen ambience.
Original prompt used in PAI (Seedance 2.5 and the variant) (2102 chars)
ShotID: rk-18-duck-v1

AUDIO
No music of any kind. Only diegetic sound: a fine spray of water, droplets ticking on dough, a violent wing shake, water sheeting over marble and glass, a startled quack and kitchen room tone. The influencer reacts with her face and body only.

LOOK, CAMERA AND PERFORMANCE
Photographic live-action cooking vlog in a bright, oversized luxury kitchen: pale-stone walls, a long white marble island, brushed brass taps, tall windows along the back wall. Strong backlight pours through those windows, rimming fur and feathers and turning flour, steam and droplets into glowing particles in the air. Fast lens wide open at T1.4–T1.8: razor-thin focus on the animal, backgrounds melt into soft bright bokeh. Human handheld camera, never still: tremors, breathing sway, imperfect framing, delayed pans and small focus corrections, held low and close to the animal. Blockbuster scale inside a kitchen: the key moment ramps into slow motion, every impact is physical and visible. The real duck keeps true anatomy, size, weight, wet white plumage and species behavior; its actual bill and wings do the spraying and the drenching. The adult influencer, in soft focus behind the island, reacts big and believable—eyes wide, a delighted grin—and treats the animal as normal cooking company.

TIMELINE — 6 SECONDS, 4 HARD CUTS
0:00–0:01.3 — Macro: the duck's bill dribbles a fine, delicate mist over the stretched dough, each droplet glowing in the backlight. SFX: a soft spray, ticking drops.
0:01.3–0:02.8 — Slow motion detail: droplets hang in the light and settle on the flour-dusted surface.
0:02.8–0:04.6 — Medium shot: the duck shakes its head and then its whole body and wings, water sheeting off in every direction, bowls and glasses hit by the spray.
0:04.6–0:06.0 — Wide shot: the entire kitchen is drenched, water running down the window, droplets on the lens; the camera jolts; the influencer, soaked, freezes wide-eyed, then bursts out laughing. SFX: quack, water everywhere.

FINAL AUDIO
No music of any kind. The last cut ends on the gag's own sound and kitchen ambience.

rk-20-peacock-v1 6 s

Utopai X Flash25 s to generate · US$ 0.79
Utopai X Flash Lite15 s to generate · US$ 0.38
Seedance 2.5 (PAI Pro)1080p · 6 s
MiniMax H3 (PAI)1080p · 6 s

Not identical: Utopai accepts up to 2,000 characters, so the opening of the look paragraph, shared by every shot, was condensed (the same way in all 13 affected shots). Audio, animal description and shot-by-shot timeline are unchanged.

Prompt sent to Utopai (1835 chars)
ShotID: rk-20-peacock-v1

AUDIO
No music of any kind. Only diegetic sound: tail feathers rattling and rustling open, steam swept sideways, a peacock cry and kitchen room tone. The influencer reacts with her face and body only.

LOOK, CAMERA AND PERFORMANCE
Photographic live-action cooking vlog in a bright luxury kitchen: pale-stone walls, a long white marble island, brass taps, tall back windows. Strong backlight rims fur and feathers and makes flour, steam and droplets glow in the air. Lens at T1.4–T1.8: razor-thin focus on the animal, soft bright bokeh behind. Handheld camera, never still: breathing sway, imperfect framing, late pans, held low and close. Blockbuster scale: the key moment ramps into slow motion; every impact is physical. The real peacock keeps true anatomy, size, weight, iridescent blue-green plumage and species behavior; its actual tail fan moves the steam. The adult influencer, in soft focus behind the island, reacts big and believable—eyes wide, a delighted grin—and treats the animal as normal cooking company.

TIMELINE — 6 SECONDS, 4 HARD CUTS
0:00–0:01.3 — Low wide shot: the peacock steps onto the island beside the steaming pizza, its folded train dragging over the marble in the backlight. SFX: claws, feathers rustling.
0:01.3–0:02.8 — Slow motion from behind the pizza: the tail rises and fans fully open, every eye-spot lighting up against the window.
0:02.8–0:04.6 — Macro: the fan sways and the steam is swept sideways in a glowing sheet over the bubbling cheese, feathers quivering.
0:04.6–0:06.0 — Medium shot: the peacock holds the full fan and turns its head toward the lens with a cry; the influencer, in soft focus behind the fan, grins and lifts a slice. SFX: cry, feathers rattling.

FINAL AUDIO
No music of any kind. The last cut ends on the gag's own sound and kitchen ambience.
Original prompt used in PAI (Seedance 2.5 and the variant) (2040 chars)
ShotID: rk-20-peacock-v1

AUDIO
No music of any kind. Only diegetic sound: tail feathers rattling and rustling open, steam swept sideways, a peacock cry and kitchen room tone. The influencer reacts with her face and body only.

LOOK, CAMERA AND PERFORMANCE
Photographic live-action cooking vlog in a bright, oversized luxury kitchen: pale-stone walls, a long white marble island, brushed brass taps, tall windows along the back wall. Strong backlight pours through those windows, rimming fur and feathers and turning flour, steam and droplets into glowing particles in the air. Fast lens wide open at T1.4–T1.8: razor-thin focus on the animal, backgrounds melt into soft bright bokeh. Human handheld camera, never still: tremors, breathing sway, imperfect framing, delayed pans and small focus corrections, held low and close to the animal. Blockbuster scale inside a kitchen: the key moment ramps into slow motion, every impact is physical and visible. The real peacock keeps true anatomy, size, weight, iridescent blue-green plumage and species behavior; its actual tail fan moves the steam. The adult influencer, in soft focus behind the island, reacts big and believable—eyes wide, a delighted grin—and treats the animal as normal cooking company.

TIMELINE — 6 SECONDS, 4 HARD CUTS
0:00–0:01.3 — Low wide shot: the peacock steps onto the island beside the steaming pizza, its folded train dragging over the marble in the backlight. SFX: claws, feathers rustling.
0:01.3–0:02.8 — Slow motion from behind the pizza: the tail rises and fans fully open, every eye-spot lighting up against the window.
0:02.8–0:04.6 — Macro: the fan sways and the steam is swept sideways in a glowing sheet over the bubbling cheese, feathers quivering.
0:04.6–0:06.0 — Medium shot: the peacock holds the full fan and turns its head toward the lens with a cry; the influencer, in soft focus behind the fan, grins and lifts a slice. SFX: cry, feathers rattling.

FINAL AUDIO
No music of any kind. The last cut ends on the gag's own sound and kitchen ambience.