Dan Kieft
Dan Kieft
300K subscribers
Subscribe
Join NextGenAI on Skool

Best AI Video Generator 2026

HERE ARE ALLMY PROMPTS

Six models, eight rounds, same prompt and same references every time. All eight prompts are on here, with the references and the clip from the round. Run them on whatever model you want and see for yourself. And if you get a different winner, come argue with me in the NextGenAI community ๐Ÿ˜Š

Tools I used

The rules

Same for every round

  • 10 seconds, 16:9 unless the prompt says otherwise, 1080p (MiniMax at 768p, Kling at 720p)
  • Same prompt, same references, every model
  • Can't generate it, or missing a feature the round needs? Instant D-tier
  • Scored S to D: absolute cinema, really good, decent, mweh, refund my credits
01

Round 1: Prompt adherence

With image references

  • Dan visits an antique shop and runs out with the sword and helmet from the references
  • Three references: Dan, the sword, the helmet
  • Five shots in 10 seconds, locked screen direction

References

Dan character sheetSword reference, two sidesHelmet reference, two angles

The prompt

@Image1 is Dan โ€” match his face, mustache, hair, white crewneck sweatshirt, grey joggers and white sneakers exactly.
@Image2 is the sword โ€” match blade runes, bronze dragon-engraved crossguard, leather-wrapped grip and lapis pommel exactly.
@Image3 is the helmet โ€” match the steel close helm with gold engravings, dragon motif, barred visor and red-and-navy feather plume exactly.
Setting: a narrow Amsterdam street, soft overcast daylight, an old antique arms shop with a dark green wooden storefront. Display window on the left two-thirds of the facade, glass door on the right third. Inside the window: the sword standing vertically on a stand beside the plumed helmet on a velvet pedestal. Photoreal, cinematic, 35mm film look, natural colors.
Screen direction:
Dan always moves left to right.
[0sโ€“2s] SHOT 1 โ€”
Medium side shot, camera parallel to the sidewalk, slow lateral track. Dan walks in from frame left, the shop window slides in from frame right. He slows, stops, turns toward the glass and leans in, eyebrows lifting with curiosity.
[2sโ€“3.5s] SHOT 2 โ€”
Over-the-shoulder from behind Dan's right shoulder, his shoulder and head soft in the left foreground. Through the glass, sharp focus on the sword and the plumed helmet in the display. His faint reflection on the window.
[3.5sโ€“5.5s] SHOT 3 โ€”
Locked-off wide shot, frontal to the shopfront, eye level, zero camera movement. Full facade in frame: display window left with the sword and helmet visible, door right, sidewalk at the bottom. Dan walks from the window to the door, pushes it open and goes inside. The door swings shut.
[5.5sโ€“7.5s] SHOT 4 โ€”
Close-up, chest-up, camera facing the shop door. The door opens and Dan steps out toward the lens wearing the helmet with the visor raised, his face and mustache fully visible, the red-and-navy plume rising above. He holds the sword upright in his right hand, blade vertical beside his face, crossguard at chest level. He breaks into a huge, genuine childlike grin.
[7.5sโ€“10s] SHOT 5 โ€”
Exact same locked-off wide shot as Shot 3, identical framing. The display window is now empty. Dan bursts out of the door and sprints out of frame right, sword raised high, plume bouncing, skipping with joy like a happy kid. Hold on the empty shopfront.
Audio:
diegetic only โ€” footsteps on pavement, distant street ambience, shop door bell chime on entering and exiting, metallic clink of the helmet, Dan's short delighted laugh as he runs off. No music.
Consistency:
same character, same outfit, same props, same storefront and lighting in every shot. No morphing, no extra people in front of the shop, helmet and sword shapes stay identical to the references.
02

Round 2: Consistency

The UGC test

  • A woman unboxes a blue wool sweater and shows the "I love AI" text on the back
  • Two references: the influencer sheet and the sweater, front and back
  • Vertical, phone front camera, three stages

References

Influencer character sheetSky-blue sweater, front and back

The prompt

[GOAL]
Vertical UGC selfie video, one young woman at home unboxing and trying on a gifted oversized sky-blue wool sweater, filmed on her own phone front camera, authentic social-media try-on, total 10 seconds.

[REFERENCES]
<The woman> corresponds to @Image 1. Use her face, skin tone, hair (dirty-blonde ponytail), makeup, earrings and the full-length dark indigo jeans. Do not use the grey background. Do not use the cap.
<The sweater> corresponds to @Image 2. Both views define ONE single sweater: oversized sky-blue chunky wool knit, ribbed crew neck, cuffs and hem, front plain, white knitted text "I love AI" across the upper back. Do not use the image background. Only one sweater exists in the video.

[CONTINUITY]
Subject: same woman throughout, ponytail, pearl drop earrings, no cap, natural glowy makeup. She wears the opaque, full-length dark indigo jeans from @Image 1 in every stage, including before and after putting on the sweater; the jeans never disappear or change into a skirt or shorts. Ordinary non-sexual everyday clothing and performance.
Scene: small bright Parisian-style apartment living room, wooden table, white wall, plant, window daylight from the side, slightly messy lived-in feel.
Prop: the same sky-blue sweater from @Image 2, never changes color, knit or text.

[STAGE 1 | 0-3s | the gift]
Initial state: handheld selfie, phone at arm's length, slight shake, close framing. She wears her black star-print bandeau top from @Image 1.
Primary event: she lifts the folded sky-blue sweater out of a kraft gift box with white tissue paper and holds it up to the lens, excited, eyebrows raised, big open smile.
End state: folded sweater fills the lower half of frame next to her face.
Cut type: HARD CUT (jump cut).

[STAGE 2 | 3-6s | wearing it, sleeve detail]
Initial state: phone now propped on the table, static front-camera framing, medium shot. She is wearing the sweater, sleeves falling over her hands.
Primary event: she leans toward the phone and brings her sleeve right up to the lens, rubbing the wool between her fingers so the chunky knit stitches and soft fuzz fill the frame, autofocus hunting briefly.
End state: extreme close-up of the blue knit texture, then she pulls back smiling.
Cut type: HARD CUT (jump cut).

[STAGE 3 | 6-10s | the back]
Initial state: same propped phone, she has stepped back, full upper body to mid-thigh in frame.
Primary event: she does one playful slow spin on the spot, sweater swinging, and stops with her back to the camera, then looks over her shoulder at the lens.
End state: her back faces camera, white knitted "I love AI" clearly readable and centered, her face in three-quarter profile over the shoulder, laughing.
Cut type: NO CUT.

[VISUAL STYLE]
Raw iPhone front-camera look, slightly wide lens distortion, natural window daylight, soft unflattened highlights, mild noise, auto-exposure shifts, no color grade, no cinematic lighting, not polished. Feels like a real TikTok/Instagram story.
[CAMERA AND PERFORMANCE]
Stage 1 handheld micro-shake, Stage 2-3 locked on a phone resting on the table, slightly low angle. Performance natural and unscripted: quick breathy laughs, talking fast, small hand gestures, genuine delight (bright eyes, hand to chest, little shoulder shimmy), not model posing.

[AUDIO]
No music. <quiet room tone, tissue paper rustling, soft wool friction near the mic, sock feet on wooden floor>.
Dialogue language: casual American English, spoken directly to camera.
Stage 1: {Okay, look what I just got as a gift... I'm obsessed.}
Stage 2: {It's so soft, look at the knit, it's so cozy.}
Stage 3, over her shoulder: {And the best part? I love AI.}
No subtitles.

[EXCLUSIONS]
No logos or brand marks, no cap, no on-screen text except the knitted "I love AI" on the sweater back, no misspelled lettering, no second sweater, no extra people, no studio lighting, no cinematic camera moves, no phone visible in frame, no watermark.

[MAINTAIN CONSISTENCY]
One woman, one sky-blue sweater, same apartment and window light across all three stages; the back text always reads exactly "I love AI".
03

Round 3: Lip sync

Desert standoff

  • Dan and Bruiser, three hard cuts, Bruiser does air quotes, Dan stays serious
  • The audio file has all three lines: Dan, Bruiser, Dan
  • Flux 3 and Gemini Omni couldn't use the audio reference, so those got generated voices

References

Dan character sheet, bald cap and aviatorsBruiser character sheet, leather jacket

The prompt

Tense desert standoff, exactly three hard-cut shots and two characters: Dan from @Image2 and robot Bruiser from @Image1. Dan is cold and serious, arms crossed and motionless except speaking lips. No smiles or smirks.

REFERENCES
@Image1: Bruiser's compact weathered ivory metal body, BLACK LEATHER JACKET, black visor, glowing blue digital eyes, antennae, joints and mechanical hands. TOTAL height soles to highest antenna tip is 0.90 m, NEVER ABOVE 1.00 m including antennae. Scale robot, jacket and hands uniformly. Helmet below Dan's belt, antenna tips around his waist. Both stand on the same ground about two metres apart. No enlarged adult-sized robot or forced perspective.
@Image2: Dan's face, mustache, THIN SILVER METAL-FRAMED SQUARE AVIATOR GLASSES WITH LIGHT BROWN-TINTED LENSES, brown denim jacket, light blue button-up shirt, gray joggers and white sneakers. Keep his visibly bad flesh-colored LATEX BALD CAP: lifted forehead seam, peeling temple patches, real hair peeking out behind. Do not smooth into a naturally bald scalp. Ignore sheet backgrounds, rulers and layouts; one character per sheet.

AUDIO
@Audio1 contains BOTH voices in three consecutive turns: DAN, BRUISER, DAN. Despite its filename it is not only Bruiser's voice. First section is Dan: "You know exactly who I am. Say my name." Middle section is Bruiser: "The AI guy who says 'this changes everything' every week?" Third section returns to the SAME Dan voice: "You're goddamn right." Match each voice to its own character. Keep every word and voice identity; tighten pauses/pace to fit. Never swap speakers.

SCENE
Flat pale yellow desert, sparse scrub, hazy mountains. Full-sized dark sedan behind Dan, black pickup behind Bruiser, nose to nose a few metres apart, characters in the gap. Harsh midday sun, short hard shadows. Warm saturated golden sand and skin against steel blue-gray/teal sky, rich contrast, fine 35mm grain. Long telephoto compression, shallow depth of field, speaking subject sharp, background blurred. Dan always looks screen left/slightly down; robot screen right/up.

SHOT 1, 0.00s to 3.20s
Clean medium close-up of Dan at eye level, right of center, little headroom, head and upper chest. Imperceptible push in. Blurred sedan front low left, soft desert horizon mid-frame, sky above. Crossed arms still. Starts immediately in low controlled menacing FIRST-SECTION DAN voice: "You know exactly who I am. Say my name." Complete line before cut.

SHOT 2, 3.20s to 7.80s
Hard cut to static reverse over Dan's shoulder. His bad bald cap, brown-jacket shoulder and some torso/belt form soft foreground at right, establishing scale. Camera at Dan's standing eye height angled gently down. Bruiser left of center in FULL-BODY view, feet planted on sand, complete antennae and BOTH hands visible throughout. Keep waist-high, 0.90 m total, never above 1.00 m. Cars anchor scale; desert/mountains blurred.
Brief nervous digital-eye hesitation, tiny weight shift, dry mechanical click; tumbleweed rolls left to right. Then hesitant questioning MIDDLE-SECTION BRUISER voice: "The AI guy who says 'this changes everything' every week?"
On "this", raise BOTH mechanical hands beside its head below antenna tips. On "changes everything", make TWO-HANDED AIR QUOTES: each index/middle finger pair curls and reopens twice simultaneously, remaining fingers curled. Lower hands on "every week?" Preserve metal fingers, no human hands. Hold shot through complete reply and gesture. Digital eyes express speech; fixed metal mouth plate. Dan stays motionless, serious and silent, lips closed.
SHOT 3, 7.80s to 10.00s
Hard cut to clean close-up of Dan from slightly low angle. Face just right of center, upper forehead cropped, chin low, only even blue-gray sky behind. Slow steady push in. Brief cold unblinking stare; fine sand veil drifts in foreground. Then quiet cold THIRD-SECTION DAN voice, same as Shot 1: "You're goddamn right." Complete all syllables and hold stern stare to end.

SOUND/CONSISTENCY
English, no subtitles/text/music/extra characters. Continuous low hot desert wind across cuts, subtle tumbleweed scratch and sand hiss. Complete breaths, clean pauses, no clipped endings, audio gates or overlapping voices.
Only current human speaker moves lips; robot uses speaker/digital eyes with fixed metal mouth plate. Preserve identities, costumes, cap, glasses, physical scale, cars, light, screen direction and three shots. Only Bruiser makes air quotes. Dan never reacts with amusement.
04

Round 4: Text-to-video

No references at all

  • A tough rapper builds a house of cards in a speeding minivan. Speed bump. Devastation
  • Found-footage look, six hard cuts, exact tattoo text
  • This is the one that shows what a model does on its own

The prompt

Raw found-footage fail video, Jackass energy. Cheap smartphone footage mixed with a 2000s consumer-camcorder look, filmed badly by unseen passengers and a roadside bystander. Unstable handheld, late whip pans, clumsy digital crash zooms, autofocus hunting, motion blur, auto-exposure pumping, blown highlights, washed-out colors and low-bitrate softness. Normal field of view, no wide-angle distortion. Bad framing, documentary realism. 16:9. 10 seconds: six hard-cut shots, ending with three seconds on his sadness.

RHYTHM:
one chronological event. Fast opening cuts, a clear exterior view through the van's bounce, then collapsing cards. Hold the final reaction without another cut. No slow introduction, slow motion or replay.

THE RAPPER:
a tough-looking, slightly overweight adult man with slicked-back black hair.
A legible black forehead tattoo reads exactly "I love House of Cards", in two lines: "I love" above "House of Cards".
Keep it separate from other tattoos.
Dense black-and-grey skull, panther, snake, dagger, rose and barbed-wire tattoos cover his body, including full sleeves, hands, knuckles, neck and upper chest.
Loose black sleeveless tank top, dark jeans, heavy gold-and-diamond chain with a large AI pendant. Identical face, tattoo wording, tattoo placement, hair, outfit and jewelry throughout. His card house means everything to him.

THE MINIVAN:
shabby, unbranded, worn grey upholstery, already speeding. Trees, fences and parked cars streak past the windows. Loud engine, road noise and cabin vibration.
He sits in the back at a low folding table secured to the floor, holding a tall, nearly finished house of ordinary playing cards. It stays upright until the bump. The van never brakes before the bump.

LOCATIONS:
cramped rear cabin and an ordinary suburban road with one thick rubber speed bump. Flat daylight, patchy blown window highlights, passing scenery shows speed.

0โ€“1s: Shot 1 (phone, shaky medium shot):
from the middle row, he leans over the nearly finished card house, intensely focused. His forehead tattoo, AI pendant and card house are visible together. Trees and fences race past behind him.

1โ€“2s: Shot 2 (camcorder look, awkward close-up):
tattooed hands and forearms as he carefully places the final card on top and withdraws his hands. The house holds.

2โ€“4.5s: Shot 3 (phone, low roadside side view):
the minivan races over the speed bump without slowing. Keep the whole van, wheels, bump and tire contact visible.
Front suspension compresses, the nose kicks up, rear wheels hit with a hard buck, and the body pitches and rebounds. The phone wobbles but keeps the van in frame. Hold through the bounce, then cut inside as that same jolt continues.
4.5โ€“6.5s: Shot 4 (phone inside, violently jolted):
the cards are already tipping and collapsing. He bounces against his seat, his chain swings and the secured table shudders. Cards scatter across the table and into the air. He lunges to save them too late. The phone jerks with the remaining bounce. The van keeps speeding. No second bump, replay or reset.

6.5โ€“7s: Shot 5 (phone, late whip pan):
messy glimpse of the last cards landing on the floor, continuing the same collapse.

7โ€“10s: Shot 6 (clumsy crash zoom into a crooked close-up):
focus finds his face quickly, then holds for three seconds. Keep his entire forehead and exact tattoo "I love House of Cards" readable.
His tough expression crumples, shoulders sag, eyes become watery and lower lip trembles as he looks down at the ruined cards. A small defeated sigh. No further pan, zoom or cut. Roadside scenery still rushes past. Cut to black at the end.

MUSIC: none.

SOUND:
raw microphone audio. Loud engine, wind, road noise, cabin rattles, front-and-rear suspension thumps from one bump crossing, brief mic distortion, fluttering cards and phone handling. A short burst of unseen passenger laughter fades before the final close-up. Vehicle noise continues beneath his sigh. No dialogue or spoken words.

EXCLUSIONS:
no subtitles, interface, timestamp, REC icon, watermark, brand logos or text overlays. Only deliberate text: physical AI pendant and exact forehead tattoo. No cinematic lighting, polished camera moves, cartoon or CGI look. No premature collapse, rebuilding card house, changing tattoos or jewelry, misspelled, mirrored, drifting or missing tattoo letters. No blur or missed pan hiding tire contact or the van's reaction. No seventh shot or extra cut during his sadness.
05

Round 5: Video-to-video

Change the boxer, the bag and the gym. Keep the punches

  • Dan in pink boxing gear, Bruiser as the punching bag, a garage instead of the gym
  • The reference video is the editing master. Movement, timing and camera stay
  • Short prompt. Say what changes, say what stays

The reference video

References

Dan in boxing attireBruiser punching bag

The prompt

Editing goal: Edit @Video 1 so Dan is boxing against the Bruiser punching bag in a simple home garage.

Source video:
@Video 1 is the sole editing master. Keep the original shot, camera movement, punches, guard, footwork, contact timing, bag swings and movement speed.

References:
@Image 1 shows the same Dan in three views. Match his face, mustache, hair, natural build, black boxing tank and boots, pink gloves and pink satin shorts. Give him a normally attached head.
@Image 2 defines the Bruiser bag: a horizontal padded capsule with off-white and orange panels, a black visor, blue dotted eyes, the 07 marking, soft antenna tabs and hanging straps. Use the subjects only, without the grey backgrounds or reference-sheet layout.

Edit scope:
Replace the boxer with Dan in the same position, wearing this outfit throughout.
Replace the red bag with the single Bruiser bag at the same size, height and suspension position. Its padding compresses at punch contact and it follows the original bag's swings.
Replace the gym setting with a modest home garage, plain concrete walls and one overhead spotlight, with the surrounding room dimmer. Preserve the framing and distance between boxer and bag.
Except for the objects explicitly modified above, all other visible characters, props, and background elements in
@Video 1 remain unchanged and are not to be replaced or removed.

Timeline inheritance:
Keep the source's complete action order and ending. Dan follows the original boxer's movements, and Bruiser follows the original bag's motion and occlusions.

Audio:
Add natural boxing sounds synchronized to the visible action: padded thuds exactly when each glove hits the bag, short exertion breaths, foot shuffles and small creaks from the swinging straps and chains. Quiet garage room tone between punches. No music or speech. -> Reference video
06

Round 6: Cinematic / realistic

The one everyone cares about

  • A woman smells a poppy, then runs through a sunlit flower field in slow motion
  • One reference: the woman and the dress
  • Five shots: close-up, tracking, macro, low angle into the sun, wide

Reference

Woman and dress reference

The prompt

[GOAL]
10-second contemplative cinematic short. A young woman gently breathes in a wildflower, then runs freely through a vast field of colorful wildflowers, seen from several angles in dreamy slow motion. Emotional arc: stillness โ†’ release โ†’ pure freedom, like a child. Total duration: 10s.

[REFERENCES]
<The Woman> corresponds to @Image 1. Use only her face, hair, makeup, and outfit: pale porcelain skin with faint freckles, warm brown eyes, flushed cheeks, red lips, shoulder-length wavy dark-brown hair with a soft side part. She wears an ivory lace corset midi dress with thin straps, a sweetheart neckline trimmed with small cream 3D roses, more cream roses scattered across a full lace skirt, a sheer tulle hem with a scalloped lace edge. Do not use the image background or the mannequin.

[CONTINUITY]
One woman only, same face, hair, makeup, and dress in every shot. Same field of wildflowers (red poppies, blue cornflowers, white daisies, yellow buttercups, tall green grass), same late-afternoon golden sun low behind her, same soft breeze. She always runs in the same screen direction, away from camera toward the sun.

[SHOT 1 | 0-2.5s | The Flower]
Tight close-up, profile, shallow depth of field: standing in the tall flowers, she raises a single red poppy to her face with both hands, closes her eyes, and slowly breathes it in. Golden backlight rims her hair and freckled skin; out-of-focus petals in the foreground. Very slow push-in.
End state: eyes closed, a soft smile beginning, the poppy against her lips.
Cut type: CUT.

[SHOT 2 | 2.5-4.5s | She Runs]
Waist-height tracking shot from behind, slow motion: she runs into the field, the camera running with her. Out-of-focus stems and petals whip past the lens in the foreground. Her lace skirt and tulle hem billow, hair bouncing, arms loose and open.
End state: her running deeper into the flowers, framed center.
Cut type: CUT.

[SHOT 3 | 4.5-6s | Touch]
Macro side tracking shot at flower height, slow motion: her open hand trails across the tops of the wildflowers as she runs, fingertips brushing petals; pollen and loose petals drift into the golden light.
Cut type: CUT.

[SHOT 4 | 6-8s | Into the Sun]
Extreme low angle from inside the flowers looking up, slow motion: she runs past above the camera against a pale blue sky, the sun directly behind her producing large rainbow lens-flare rings and a soft halo. Her dark hair streams across the frame, her face tilted up, eyes closed, laughing silently.
Cut type: CUT.

[SHOT 5 | 8-10s | Free]
Wide static shot, slow motion: the immense flower field under a pastel sky, she is small in the frame, running away toward the low golden sun, arms spread open, her ivory dress glowing in the backlight, slowly swallowed by flowers and light.
End state: a tiny ivory figure dissolving into the golden horizon.

[VISUAL STYLE]
Dreamy, poetic indie-film aesthetic shot on 35mm: soft pastel blue sky, warm golden backlight, saturated but soft summer greens and flower colors, the ivory lace glowing translucent in the sun, natural lens flare, gentle halation, fine film grain, slightly lifted blacks, warm skin tones. Photorealistic. Calm, nostalgic, contemplative.

[CAMERA AND PERFORMANCE]
Everything in graceful slow motion; floating, unhurried camera even while tracking. Her emotion is shown through cues, not acting: closed eyes and a slow inhale (stillness), relaxed shoulders and open arms (release), a wide carefree smile and streaming hair (freedom).

[AUDIO]
(A soft, emotional piano melody with airy strings that swells gradually into a warm, uplifting crescendo on shots 4-5.)
<A gentle inhale in shot 1, soft wind, rustling grass and flowers, light footsteps, distant birdsong, her soft breathless laughter.>
No dialogue. No subtitles.

[EXCLUSIONS]
No other people, no text, logos, or watermarks. No buildings, roads, or fences in frame. No duplicate woman. No mannequin.

[MAINTAIN CONSISTENCY]
One woman identical to @Image 1 in every shot; same ivory lace rose dress, same flower field, same golden late-afternoon light, consistent running direction toward the sun.
07

Round 7: Animation

Pixar-style

  • A cat tries to pop the duck's pool float. It backfires. The duck ends up on the cat's head
  • No references, one continuous shot, three stages
  • The continuity lock does the heavy lifting here

The prompt

Goal:
A scheming black cat, safe and dry on the pool edge, punctures the flamingo float the duck is lounging on to dunk it. To reach it he braces one front paw on the float, so when it deflates it takes him down too. The duck floats happily without ever noticing, and the cat ends up as the duck's new pool float.

Continuity lock:
One male black cat with green eyes, one small cheerful yellow duck about the size of the cat's head, and one pink inflatable ride-on flamingo pool float. The flamingo is an inanimate vinyl prop with a solid, rounded body and no hole or opening. It is about twice the cat's length, its tail end is closest to the cat, and its curved neck and head rise from the far end. In a sunny backyard pool, the cat crouches dry on the tiled pool edge on the left. The flamingo floats just beyond his reach on the right, and the duck sits on top of the flamingo's back, in the middle between tail and neck. Both characters are dry and the flamingo is fully inflated at the start.

Visual style:
Pixar-style 3D animation with expressive faces, soft fur, glossy vinyl, turquoise water and warm summer light. Simple backyard surroundings keep attention on the characters. Playful slapstick with believable weight, water and material changes.

Stage 1, the scheme:
One continuous side shot at water level, with the cat on the edge, the flamingo with the duck on its back, and the gap of water between them clearly visible. The float gently bobs. The duck sunbathes on the flamingo's back with its eyes closed and a blissful little smile.
The cat glances at the duck to make sure it isn't watching, narrows his eyes and gives a wicked grin.
He stretches out over the water, hind paws on the tiles and one front paw braced on the flamingo's tail end, his body bridging the gap.
With his other front paw he slowly unsheathes one claw, which catches the sun with a tiny glint, and presses it into the side of the flamingo's body near the tail. The float stays fully inflated until the claw touches it.

Stage 2, the backfire:
Same shot, no cut. The claw pierces the vinyl and air hisses out of the small hole into the cat's face, ruffling his whiskers. The flamingo's body starts to wrinkle and sag, and its neck begins to droop.
The float tilts away from the cat and the duck slides gently off its back into the water, where it keeps floating with its eyes still closed, completely untroubled.
The cat grins in triumph for one beat.
Then the softening vinyl under his braced paw caves in and the float drifts away from the edge. His body stretches into a long, wobbling bridge, his grin frozen, until his hind paws slip off the tiles and he drops face-first into the water with a splash. The flamingo keeps deflating until its neck flops onto the water.

Stage 3, the payoff:
Still no cut. The camera slowly pushes in to a medium close-up of the cat as he surfaces and treads water with small paddling strokes, his head above water. His fur is plastered flat, ears pinned back, water dripping from his whiskers.
A short distance behind him on the right, the flamingo lies completely flat on the water, an empty, wrinkled pink skin with its neck and head collapsed, and it stays flat until the end.
The duck paddles over, flutters up onto the top of his head and settles down with its eyes closed, using him as its new float.
Under the duck's weight the cat sinks slowly until the water sits just below his eyes. He holds completely still, then slides his eyes to the camera with a flat, miserable stare. One bubble rises in front of his nose and pops. The duck gives one content quack.

Audio:
<Gentle pool water lapping> <A tiny sparkle ting as the claw glints> <A sharp air hiss starting at claw contact> <A small plop as the duck slides in> <A rubbery squeak as his paw slides on the wet vinyl> <A splash as the cat drops in>
<A soft sputter as he surfaces> <Quiet dripping> <One bubble blub> <One content duck quack>. Wordless comedy, no music, no narration.

Final consistency:
Preserve both character designs, their relative scale and the flamingo's pink color and shape.
There is only one flamingo float, and its state only moves forward: fully inflated, then sagging, then completely flat. Once flat, it never reinflates.
The duck sits on the flamingo's back at the start and is in the water after it slides off. The cat changes from dry and smug on the pool edge to soaked and miserable in the water, and he never touches the flamingo after falling in. The duck stays cheerful, relaxed and unbothered throughout and never panics.
08

Round 8: Motion graphics

Kinetic typography

  • Five phrases, bold text and hands moving through connected layouts, one illustrated eye at the start
  • Every phrase has to read clearly, the last one has to hold
  • Exact colours, exact spelling, exact order

The prompt

GOAL:
Create a dense, fluid kinetic typography sequence built around five short phrases: LOOK CLOSER, THINK BIGGER, MAKE IT DIFFERENT, MORE IDEAS and CREATE WITHOUT LIMITS. Animate the words within each phrase separately, letting one word's movement trigger the next. Hands and letter shapes guide the transitions. Finish on a complete CREATE WITHOUT LIMITS poster.

DESIGN:
Ultra premium 2D motion graphics with the precision and finish of an After Effects piece made by a top motion design studio. Dark psychological thriller mood, punk zine collage meets Swiss editorial design. Dark red (#B80000), ink black (#0A0A0A) and paper white (#E6E6E2). Massive condensed capital letters, posterized halftone hands and subtle print grain. Strong asymmetric poster compositions and deliberate negative space. All text is large and bold. Typography carries the action, with hands integrated into its movement.

MOTION AND CAMERA:
One continuous virtual camera journey with no hard cuts. Connect the layouts through letter-counter zooms, rotating letter masks, moving bars and matched shapes. Use quick anticipation, decisive acceleration, fluid follow-through and controlled overshoot. Keep foreground lettering, background outline echoes and graphic accents moving at different speeds. Words land with brief readable beats, while secondary layers continue moving. The final phrase receives the longest hold.

STAGE 1, LOOK CLOSER: On red, white LOOK slides in horizontally. One small illustrated eye, a flat white almond with a solid black dot, appears in the final O. The first O stays an ordinary letter. A halftone fingertip taps the final O; the icon blinks once and folds away. CLOSER rises beneath LOOK, letter by letter, drawing the rows into one bold poster. With the icon gone, the camera dives through the O's empty counter into the next layout.

STAGE 2, THINK BIGGER: On paper white, black THINK arrives from above while red BIGGER slides in from the side beneath it. A black underline sweeps under THINK, sending a wave through its letters. BIGGER expands, pushing THINK towards the top edge. Large outline echoes follow the expansion with a slight delay. The camera rolls a quarter-turn through the I stroke; that stroke becomes a diagonal mask revealing black.

STAGE 3, MAKE IT DIFFERENT: White MAKE sweeps across the upper part of the black composition, IT drops into the centre, and red DIFFERENT assembles across the bottom from alternating directions. Two halftone hands pull the upper and lower rows sideways in opposite directions. The solid word shapes remain recognisable while outline echoes stream behind them. IT pivots between the moving rows like a hinge. The hands sweep the rows into broad moving bars that reveal the next layout.

STAGE 4, MORE IDEAS: On red, huge white MORE expands towards the lens. IDEAS passes behind it as an oversized black outline, then becomes solid white as MORE moves upward. The hands glide beneath IDEAS and lift the whole composition towards the camera. Its expanding white letter strokes fill the image and become the next background.

STAGE 5, CREATE WITHOUT LIMITS: On paper white, black CREATE drops into the upper row, red WITHOUT slides into the middle, and black LIMITS rises into the lower row. Each word lands on a separate percussion accent. A thick red bar runs behind LIMITS while the hands form a frame at opposite corners. The camera returns level and eases back as the letters settle. Hold the full CREATE WITHOUT LIMITS phrase large, sharp and readable.

AUDIO: A driving electronic rhythm with dry percussion, deep bass and rising synthetic sweeps. Tactile clicks on taps and word landings, layered whooshes following the type rows, a rising sweep during the letter wave and a deep final impact. Sound accents follow individual word movements. No speech or narration.

CONSISTENCY:
Preserve the type family, hand designs and three-colour palette. The single illustrated icon stays flat, with no realistic anatomy or glossy shading. After its opening blink it disappears and never returns. Later layouts use typography, hands, bars and shapes. Use the exact phrase order and spelling: LOOK CLOSER, THINK BIGGER, MAKE IT DIFFERENT, MORE IDEAS, CREATE WITHOUT LIMITS. Keep each complete phrase readable at its landing. Transitions visibly carry the movement between layouts. The final phrase remains stable and complete.
Copied