Sticker Fashion Transformations
A 30-second vertical fashion editorial uses four fixed outfit stickers, a mouse cursor, and full-body fly-in overlaps to transform one consistent protagonist through five K-pop-inspired looks.
99 prompts
Text-to-video, image-to-video, camera control, brand films, and video model evaluation prompts.
A 30-second vertical fashion editorial uses four fixed outfit stickers, a mouse cursor, and full-body fly-in overlaps to transform one consistent protagonist through five K-pop-inspired looks.
A timecoded 30-second rooftop sequence follows a runner from a static establishing shot into a hip-height tracking move and a final arc against the illuminated city skyline.
A three-scene martial-arts fantasy comparison combines a folding cyber-inn mirror dimension, a snow-canyon transition, and a desert monster battle using three image references and match-cut editing.
A 30-second high-budget aerial combat sequence uses five-second timeline blocks, aggressive camera changes, speed ramps, tracer fire, missile runs, and a final golden-hour climb above the clouds.
A timecoded 30-second medieval procession escalates from restrained public condemnation to an overwhelming wall of bells and chants while the central noblewoman's composure gradually breaks.
A timecoded 30-second action sequence keeps a blonde fighter in a fixed tactical outfit while she chains acrobatic katana attacks through monsters and a dark-knight boss.
A 30-second fantasy-comedy one-take follows an ancient forest spirit as he transforms into a man, joins a campsite, and delivers a mock-serious beer-commercial payoff.
A 15-second vertical food commercial keeps four family members consistent while coordinating macro product shots, Mandarin narration, character dialogue, and precise lip sync.
A single continuous tracking shot transforms a falling ink drop into living calligraphy, liquid ribbons, a floating dragon, and neon graffiti before swallowing the camera.
A high-energy urban fashion film follows a model through New York with an ultra-wide fisheye lens, extreme low angles, whip pans, jump cuts, light trails, and music-player overlays.
A 30-second vertical one-take presents ancient Rome in 80 AD as raw, unstable smartphone footage, passing a market, slave auction, tavern brawl, legionaries, and the Colosseum.
A handmade clay mug waits behind perfect porcelain cups in a 30-second stylized 3D short, using three image references, tightly locked character animation, and a 13-segment emotional storyboard.
The same 15-second one-take desert-rider prompt is tested across Seedance 2.5, Seedance 2.0, and MiniMax H3, with a fixed lateral tracking camera and evolving background action.
A 20-second luxury fragrance film uses an Omni Reference portrait and a six-part timestamped storyboard to control identity, pacing, camera movement, product reveal, and dialogue.
A 30-second premium beer spot uses rapid timecoded cuts, kinetic camera movement, macro product shots, desert-road energy, and a final product hero reveal.
An unseen housemate washes dishes, cooks dinner, steals a fry, tucks its exhausted roommate in, and finishes the night with a cold beer in a cozy 30-second comedy short.
An elven princess defeats 36 visually distinct orcs through precisely counted flash relocations and sword contacts in a tightly choreographed 30-second continuous pursuit shot.
A face-referenced rider races through volcanic ash on an obsidian dragon, ignites a sword during a slow-motion crash zoom, and sends a giant leviathan into erupting calderas.
A face-referenced rider skims a storm-tossed ocean on a bioluminescent dragon, draws an igniting sword in slow motion, and cleaves a colossal leviathan in one continuous shot.
Two performers move through pink studios, blue spotlights, tinsel backdrops, glamour close-ups, and synchronized choreography in a tightly timecoded 30-second K-pop music video.
A chest-mounted single-take POV follows a sequence of skyscraper-edge pole jumps, including a near-fatal slip, with visceral wind, breathing, heartbeat, and metallic impacts.
A tense first-person alley confrontation uses a continuous handheld POV, tightly timed Indonesian dialogue, accurate lip sync, and nonviolent physical blocking.
The complete fable of the proud hare and steady tortoise is compressed into a colorful 30-second family-oriented 3D animation.
A performer rises inside a late-1990s desktop paint interface, draws a glowing five-point star around himself, and exits as the camera appears to pass through the monitor.