Sign up and save up to 50% on a yearly plan.Sign up free →
Pricing, save up to 50% on yearly plans

AI video effects. Freeze the park. Catch the drop.

Upload one photo. Pick an effect. Star in the moment.

46 AI video effect workflows

Turn to Stone

Face down a charging attacker and turn them to stone with a curse, ending on a close-up of their petrified face. Upload TWO photos: the spellcaster, and the one who gets cursed. 15s.

1 step
Seedance 2.515s · 9:16
Inspired by OpenClipson organic

DISSOLVE — Sand & Reform @image1 dissolves into a swirling ribbon of glowing golden sand that sweeps sideways through the air and knits back together a few steps away, reforming in a new pose — a clean, magical teleport-style dissolve. Cinematic 15s, no ash, no gore. Swap in ANY avatar.

1 step10
Seedance 2.515s · 9:16
Inspired by OpenClipson organic

FLOWER MORPH — Petal Explosion Reveal @image1 claps and bursts into a huge cloud of colourful petals that fill the frame — then the petals rush back and reassemble them in a stunning floral look (a flowing gown or a sculpted floral suit, whichever suits their gender presentation) for a confident final pose. Punchy, magical, cinematic 15s. Swap in ANY avatar.

1 step10
Seedance 2.515s · 9:16
Inspired by OpenClipson organic

BEAST TO HUMAN — Awakening A majestic lion gathers golden light, its fur dissolving into glowing embers, and from the light @image1 steps forward as a clothed human. A clean, magical beast-to-human transformation in a moonlit stone temple, one cinematic 15s take. Swap in ANY avatar.

1 step
Seedance 2.515s · 9:16
Inspired by OpenClipson organic

Your avatar turns to stone from the feet up, aware and struggling, ending as a weathered statue among the rocks. Music only, loud. See the video step for the full prompt.

1 step
Seedance 2.515s · 9:16
Inspired by OpenClipson organic

FIRE VILLAIN — Ignition Transformation Audio: music only, mixed loud and full so the music dominates the entire soundtrack from the first frame: one continuous dark epic cinematic instrumental — low ominous strings, a slow rising tension build, deep percussive hits, and a huge climactic swell held long on the final villain reveal. NO background or ambient sound of any kind, no footsteps, no foot taps or scuff, no foley, no fire crackle competing with the music, no recognisable copyrighted melody, no spoken dialogue, no invented language. IDENTITY LOCK — @image1: keep the exact face, bone structure, eyes, eyebrows, nose, mouth, skin tone, hairline, hairstyle, hair length, texture and colour, freckles or marks, apparent age, build and proportions identical throughout — the SAME recognisable person the whole time. The attached @image1 outranks every word: where text and image disagree, follow the image. @image1 is an adult, the only person in the frame, and never turns into anyone else. One single being throughout. WARDROBE — @image1 is fully clothed and decent at all times in a fitted modern outfit taken from @image1, torso and legs covered, never barefoot, never undressed, no exposed underwear, nothing revealing. The clothing chars and smoulders at the edges as the flames rise but stays on and never burns away or exposes the body. THE TRANSFORMATION — a dramatic cinematic comic-book SUPERVILLAIN ignition in which @image1 is UNHARMED and POWERED by fire (a fire-wielding super-villain), never injured, never in pain. It builds slowly for real drama. At first @image1 does NOT know what is happening: standing normally, then a faint flicker of embers at the fingertips makes @image1 look down, confused and uneasy. Then the HANDS suddenly burst into roaring flames — @image1 recoils in shock, staring wide-eyed at the burning hands, staggering back. The fire races up the arms, over the shoulders and across the whole body, wreathing @image1 in big roaring flames, flying embers and heat-haze — unhurt, the skin glowing with inner fire and cracks of molten orange light, the eyes igniting to a burning glow. The fear slowly turns into dark, confident, menacing power. THE CLIMAX (long, dramatic): fully wreathed in flame, @image1 slowly straightens to full height, lifts the head, and a cruel evil grin spreads across the fire-lit face; then @image1 SPREADS both arms wide and the flames FLARE outward and upward in a huge burst, roaring higher, embers exploding around the body — @image1 holds this powerful menacing villain pose, wreathed in towering flames with glowing ember eyes, radiating danger, held long to the very last frame. STAGE TIMING — one direction only, always forward, with the climax given the most time: unaware with the first flicker, then the hands ignite and the shocked recoil, then fire spreading up the body as fear builds, then fully wreathed and the shift into dark power, then the big villain climax. Never reversing. TIMELINE (20 s): 0–4s: @image1 stands normally on the night street, then a faint flicker of embers dances at the fingertips; @image1 looks down, confused and uneasy; the camera holds a medium. 4–7s: the hands suddenly burst into roaring flames; @image1 recoils in shock, staggering back, staring at the burning hands; the camera pushes in to the hands then tilts to the face. 7–11s: the fire races up the arms, shoulders and torso, embers flying, the skin glowing with molten orange cracks; @image1's fear starts turning to dark resolve; the camera eases back and slowly orbits. 11–15s: @image1 is fully wreathed in flame, the eyes igniting to a burning glow, the frightened look hardening into menacing power; the camera settles low and wide. 15–20s: the big villain climax: @image1 straightens to full height, a cruel evil grin spreading, then SPREADS both arms wide as the flames flare huge and roar upward, embers exploding around the body, holding the towering menacing villain pose with glowing ember eyes, held to the very last frame. DIRECTION — one continuous handheld-on-gimbal take: open on a medium of @image1, push in to the hands as they burst into flame, tilt to the shocked face, ease back and slowly orbit as the fire wreathes the body, then crane back and slightly low for the final towering villain stance as the flames flare wide. Continuous screen direction; no cuts, no teleporting, no extreme shake. SCENE — a dark city street or rooftop at night: wet reflective asphalt, distant city lights and neon, deep shadows, drifting smoke, so the fire glows brightly and dramatically against the darkness and reflects on the ground. A real night city setting with depth — never a plain grey or black empty studio void. Let each beat play out fully and finish its action before the next; ONE continuous coherent take, no cuts and no scene changes. CHOREOGRAPHY — believable reaction and weight: the uneasy look down, the recoil as the hands ignite, the stagger, the slow shift from fear to villainous power, then the big deliberate arms-wide flare on the climax; real fire behaviour with flames rising, embers drifting and heat-haze; escalating, never idle. ANATOMY — @image1 has exactly two arms and two hands with five fingers per hand, two legs and two feet; unharmed and whole throughout, no gore, no blood, no burned or injured flesh, no charring of the body — just glowing fire power over an intact, fully clothed person. LOOK — ultra-photorealistic cinematic, 50mm close-up and 35mm wide, shallow depth of field, real roaring fire, embers, smoke and heat-haze, glowing orange rim light on the skin, dark night grade with strong fire glow, realistic grain, natural motion blur, cinematic shutter. Never plastic, never cartoon, never CGI-looking. NEGATIVE — the person looking hurt, injured, in pain, burned, charred, with burnt or blistered skin, wounds, gore or blood; any sense of self-harm or suffering; undressing, the clothing burning away, a bare torso, exposed underwear or revealing clothing; the fire not actually appearing; only a small flame instead of full engulfment; a weak or rushed climax; the face becoming unrecognisable or changing identity; a plain grey or black empty studio void instead of the night city; the change reversing; a warped face, taffy limbs, wrong proportions, body distortion; feet sliding or gait glitching; the camera cutting, spinning or teleporting; a scene cutting off mid-action; jumping to a random unrelated scene before the current action finishes; skipped or truncated beats; no character vanishing, dissolving, fading or teleporting into thin air — every exit is physical; @image1 duplicating; a second person; a third hand, extra arm, more than two hands, extra or missing fingers or limbs; slow motion, speed ramps, repeated loops, aerial shot; anime, cartoon or CGI look; on-screen text, captions, numbers, readable signage, invented logos or watermark; any background, ambient or foley sound, footsteps or fire crackle competing with the music, music too quiet or not clearly audible; only the elements explicitly described appear.

1 step
Seedance 2.520s · 9:16
Inspired by OpenClipson organic

Your avatar melts like a candle, aware and reacting, until the whole body runs out into a pool of wax on the floor of their room. Music only, loud. See the video step for the full prompt.

1 step
Seedance 2.510s · 9:16
Inspired by @Noor_ul_ain43on X

Generate a video where the garment in the boutique window flies through the glass and becomes what you are wearing.

2 steps601.7K
Seedance 2.515s · 9:16
OpenClips

MELT — LIQUID CHROME Cornered in a rain-slicked neon alley, @image1 melts into mirror-bright liquid chrome, pours away to escape, then re-solidifies into themselves. Plug in ANY avatar. Music only, loud. See the video step for the full prompt.

1 step
Seedance 2.515s · 9:16
Inspired by OpenClipson organic

TORNADO — Summon & Hurl @image1 pulls a small swirling vortex of wind into their hands, compresses it into a glowing storm-orb and hurls it toward the horizon — where it erupts into a colossal towering tornado. A dynamic, powerful cinematic 15s take. Swap in ANY avatar.

1 step
Seedance 2.515s · 9:16
Inspired by OpenClipson organic

FLOWER MORPH — The Bush Becomes a Person @image1 walks up and touches a big flowering bush — and the blossoms spiral upward and knit into @image2, who steps out in flowing floral couture (a gown or a flowing tailored look, whichever suits their gender presentation). A magical two-avatar transformation, cinematic 15s. Plug in ANY two avatars.

1 step
Seedance 2.515s · 9:16
Inspired by OpenClipson organic

LIQUID SHIFT — Silver Midas @image1 turns to gleaming liquid silver — clothes and all — then reaches out and touches an apple, and it turns to shining silver too. A clean, elegant Midas-touch transformation, one continuous cinematic 15s take ending on an extreme close-up of the metallic face. Swap in ANY avatar.

1 step
Seedance 2.515s · 9:16
Inspired by OpenClipson organic

Your avatar blooms into a lush red-flowering garden bush — head becoming one big red flower, aware and reacting, until no skin is left and the figure reads as a real leafy shrub of petals and green foliage blending into the hedges. Music only, loud. See the video step for the full prompt.

1 step
Seedance 2.515s · 9:16
Inspired by OpenClipson organic

Your avatar turns to ash and is carried off on the wind, aware and reacting with a dramatic raised hand, in a windswept field at golden hour. Music only, loud. See the video step for the full prompt.

1 step
Seedance 2.515s · 9:16
Inspired by OpenClipson organic

Your avatar transforms into the animal from the reference image in one dramatic cinematic take, close on the face. Music only, loud. See the video step for the full prompt.

2 steps
Seedance 2.510s · 9:16
Inspired by @meAsifAion X

10 seconds, 16:9, five shots, bright music. @image_1 films a selfie vlog through a pastel-pink shopping lane and the phone filter draws pink augmented-reality effects over the picture. @image_1 is the only attached asset and the same person in every shot. ATTACHED IMAGES ARE MANDATORY AND FINAL: @image_1 decides the face, hair, skin and build in every frame. Where this text and the attached image disagree, the attached image wins. Preserve the gender presentation of @image_1 exactly as shown. Every hand in frame has five separate fingers. WARDROBE LOCK: one outfit for the whole film - one simple plain everyday top in a single solid colour, the hair exactly as @image_1 already has it, at whatever length and texture. The garment is fixed; only its drape adapts to the body of @image_1. THE PLACE: a narrow pedestrian shopping lane on a bright overcast day, storefronts in pastel pink and cream, café tables and folding chairs out front, chrome balls and pink parasols along the paving, Korean shop signage on the fascias, soft even daylight. CAMERA: phone front camera at arm's length on a very wide, slightly barrel-distorted lens. The first two shots sit inside a soft round vignette; from the third the frame opens to the full rectangle. Natural phone colour, handheld drift, focus and exposure a beat late. THE EFFECTS: pink augmented-reality graphics drawn over the live picture the way a phone filter lays them on - a ribbon of light that writes itself through the air, small chrome hearts turning with real reflections, little square photo stickers with soft white borders, and one large glowing heart outline. They sit flat on top of the image, in front of everything including the arm, cast no light on the face and never hide it. [0-2s] Already filming. @image_1 looks into the lens and starts talking, one hand up near the collarbone, inside the round vignette. [2-4s] @image_1 sweeps one hand through the air and a ribbon of pink light follows it and writes a heart that hangs there, sparkling, before it fades. [4-6s] The frame opens out. Two chrome hearts and two photo stickers drift in, turning slowly. @image_1 winks and brings both hands up beside the jaw. [6-8s] Cut wide, the camera no longer in their hand: @image_1 walks away down the middle of the pink lane and turns back over one shoulder. [8-10s] Back to the selfie. A large glowing pink heart outline draws itself around the head and shoulders of @image_1, who makes a small finger heart close to the lens and laughs. A bright light instrumental at one even level with street ambience under it; @image_1 talks brightly to the lens throughout, the words under the music rather than over it. The five shots run back to back with no gaps and no black frames; the first frame is already lit and already moving. Writing in frame is only the shop signage that belongs to the place. No subtitles.

1 step973.2K
Seedance 2.510s · 1:1
Inspired by OpenClipson organic

DISSOLVE — Butterfly @image1 stands serene, then gently breaks apart into a swarm of glowing butterflies that scatter toward the camera and sweep up into the sky. A magical, beautiful cinematic 15s dissolve — no ash, no gore. Swap in ANY avatar.

1 step
Seedance 2.515s · 9:16
Inspired by OpenClipson organic

SANDSTORM SUMMON — Desert Vortex @image1 summons a colossal sandstorm in the desert — the camera sweeps 180° from their face around to behind them, revealing a towering vortex rising far ahead, which they conjure and then collapse with a clench of the fist. Epic, controlled, cinematic 15s. Swap in ANY avatar.

1 step
Seedance 2.515s · 9:16
ai_lifehack55

15 seconds, 1:1. Photoreal night Shibuya. Hard cuts. Cinematic. PERSON LOCK: @image1 is the person on the STREET — photocopy their FACE, hair and features exactly as in @image1. Preserve their gender presentation exactly as shown. Use they / their / them only. IGNORE the background of @image1. STREET OUTFIT: exactly the clothes worn in @image2. Same person in the crowd shots. POSTER LOCK: @image2 is the ONLY billboard ad — the person exactly as pictured in @image2. Photocopy that poster onto the giant screen. Do not redesign. Night city, crowds glance and point. The person walks, confused, looks UP. 0.00–1.70 High angle sidewalk, the person walking. HARD CUT. 1.70–3.30 Side follow. Passerby turns, someone points. HARD CUT. 3.30–4.90 Front face CU, slight confusion. HARD CUT. 4.90–6.60 High oblique, more looks. HARD CUT. 6.60–8.50 Follow. The person checks left/right, then looks up. HARD CUT. 8.50–10.80 Face closer. Gaze UP. HARD CUT. 10.80–15.00 Building-wall digital billboard showing @image2. FIRST the full @image2 poster fills the screen. Camera PUSHES IN continuously to a CLOSE-UP of their face. Then the person ON THE POSTER winks once and freezes. HOLD. AUDIO: city, cinematic BGM. No speech. NEGATIVE: character-sheet text, overlay captions, the @image1 background, extra hero faces.

1 step
Seedance 2.515s · 1:1
Inspired by OpenClipson organic

ANIMAL MORPH — Into a Lion @image1 sprints and leaps, and in cinematic slow-motion detail transforms mid-air into a powerful lion — skin rippling into fur, a hand curling into a clawed paw, the face becoming a lion's — then lands hard and roars. A film-style werewolf-movie morph, one continuous 15s take. Swap in ANY avatar.

1 step
Seedance 2.515s · 9:16
Inspired by OpenClipson organic

FIRE — Phoenix Embers @image1 ignites into a controlled swirl of fire and glowing embers that rise into the shape of a great phoenix, then cascade back down and reform @image1 standing calm and unharmed — on a dark volcanic plain. Music only. One continuous, cinematic 15s take. Swap in ANY avatar.

1 step
Seedance 2.515s · 9:16
Inspired by OpenClipson organic

Your avatar turns into fully transparent, see-through water and then suddenly collapses and splashes down into a clear puddle on a sunny poolside. Music only, loud. See the video step for the full prompt.

1 step
Seedance 2.515s · 9:16
Inspired by OpenClipson organic

TORNADO IN THE CITY — Pulled Into the Vortex Audio: one continuous dramatic cinematic instrumental, studio-recorded, loud and forward — low pulsing strings, a deep sub-bass rumble, a rising tension build and heavy percussive hits on the biggest moments. The picture itself has no recorded sound. IDENTITY LOCK — @image1 is the one person we follow: keep the exact face, bone structure, eyes, eyebrows, nose, mouth, skin tone, hairline, hairstyle, hair length, texture and colour, freckles or marks, apparent age, build and proportions identical in every frame. The attached @image1 outranks every word: where text and image disagree, follow the image. @image1 is an adult and never turns into anyone else. Other people are anonymous background civilians whose faces are never featured; they never match or duplicate @image1. WARDROBE — @image1 wears an everyday outfit taken from the reference and unchanged; fully dressed, never barefoot. THE SEQUENCE — a photoreal disaster sequence in a big city during a tornado, following @image1 who FIGHTS the storm rather than waiting for it: bracing, gripping, being dragged, and finally torn off the ground and pulled INTO the funnel, ending inside the spinning vortex where @image1 tumbles and screams in terror. The tornado must physically advance ALL THE WAY down the street and engulf @image1 up close — never merely reach from a great distance. DIRECTION — one continuous handheld storm-chaser take: wide establishing down the avenue, running follows, low angles as the funnel bears down, then a tight handheld ride as @image1 is dragged and lifted, and finally an interior vortex shot circling @image1's terrified face. No hard cuts; one continuous shot; no teleporting. SCENE — a wide city avenue between skyscrapers with the storm sky above, wet reflective road, cars and debris, then the churning grey-brown interior of the tornado full of swirling dust, papers and debris for the final beats. Nobody's face is featured except @image1. TIMELINE (30 s): 0–4s: wide down the avenue: skyscrapers and a huge tornado funnel on the skyline, overcast storm light, wet street, crowds of civilians fleeing past; @image1 stands their ground, bracing against the wind, clothes and hair snapping. 4–8s: the funnel visibly ADVANCES down the street toward @image1, growing fast; @image1 grabs a lamp post and holds on, leaning hard against the pull, feet sliding on the wet asphalt as the wind drags at them; debris flies past. 8–13s: the tornado reaches them: it fills the street right around @image1, the roar peaking; @image1 clings to the post with both hands, legs lifting off the ground, body streaming horizontally in the wind, fighting with everything, face straining — a wide-to-medium shot, no close-up yet. 13–17s: the grip is torn loose; @image1 is ripped off the post and dragged low across the ground, clawing at the asphalt, then snatched upward as the funnel takes them; the camera cranes up following the body as it is pulled into the base of the vortex. 17–23s: INSIDE THE TORNADO: @image1 tumbles and spins inside the churning grey-brown column, surrounded by swirling debris and dust, limbs flailing against the spin, hair and clothes whipping, rising through the vortex. 23–30s: CLOSE-UP inside the vortex on @image1's terrified face: eyes wide, mouth open screaming, hair lashing across the face, debris streaking past, the body spinning; hold on the terror as @image1 is carried higher into the roaring funnel to the last frame. CHOREOGRAPHY — believable resistance and weight throughout: bracing, gripping the post, feet sliding, being torn loose, clawing, then the helpless tumble and scream inside the vortex; real inertia and violent wind on hair, clothes and debris; escalating, never idle, never passively waiting. ANATOMY — @image1 has exactly two arms and two hands with five fingers per hand, two legs and two feet; dramatic but never grotesque, no gore, no blood, no dismemberment. LOOK — ultra-photorealistic disaster-film realism, 28mm and 35mm, deep focus on the street and funnel then tight on the face in the vortex, cool desaturated storm grade, rain sheen, heavy volumetric dust and debris, realistic grain, strong natural motion blur, cinematic shutter. Never plastic, never cartoon, never CGI-looking. NEGATIVE — @image1 passively standing and waiting to be sucked up instead of resisting; the tornado only reaching from a great distance without advancing close and engulfing them; no interior-of-the-tornado shot; no close-up of the screaming terrified face; the featured person's face changing or duplicating; another person's face featured or matching @image1; a costume or props instead of real storm effects; the lift snapping instantly with no struggle; the body dismembering, gore, blood; a stretched or warped face, taffy limbs, wrong proportions, body distortion; feet sliding unnaturally or gait glitching; hard cuts, the camera teleporting, jumping to an unrelated scene; a scene cutting off mid-action; skipped beats; slow motion, speed ramps, repeated loops, aerial establishing drone shot; plastic or beauty-filtered skin, anime, cartoon or CGI look; on-screen text, captions, subtitles, numbers, readable signage, invented logos or watermark; only the elements explicitly described appear.

1 step
Seedance 2.530s · 16:9
Inspired by OpenClipson organic

Photoreal cinematic street scene, premium commercial grade, sunny palm-lined luxury shopping street, shallow depth of field, natural daylight, subtle filmic grain, realistic skin, fabric and liquid — an accidental sidewalk bump that spills a coffee and knocks the sunglasses off, played spontaneous and glossy, building into two hyper-zoom ultra-slow-motion brand-name hero shots, never aggressive. LEAD is exactly the person in @image1 — preserve their exact face and identity throughout, real skin, exactly two hands. Preserve their gender presentation exactly as shown. Use they / their / them only. They hold a smartphone in the LEFT hand pressed against the LEFT ear, clearly mid phone call to someone else, hand visibly gripping the phone at ear level (never tucked under the arm or armpit). In the RIGHT hand they hold a coffee cup exactly like @image2 (same cup, colours and readable brand logo exactly from the reference). Sunglasses exactly like @image3 (same frame and readable brand logo exactly as on the reference) worn on the head. They walk relaxed, smiling, talking into the phone, saying only: "Hello?". A SECOND PERSON with a dog enters diagonally from the right on their own independent path that simply happens to intersect — clearly a different-looking person, not resembling @image1. The second person glances at the dog; @image1 is on the call. Neither targets the other, neither accelerates — no lunge, squared shoulder, push or body-check. BEATS (one continuous take; real-time, then hyper-zoom ultra slow motion on the brand names, then real-time): 0-2.5s camera tracks BACKWARD at heel height then tilts up to a medium of @image1 walking toward us, the phone in the left hand at the ear mid-call, the @image2 cup in the right hand, the @image3 sunglasses catching sun. 2.5-4.5s the second person and the dog cross in from the right on their own line, paths converging naturally. 4.5-6s they notice each other at the last instant, both twist to avoid contact, shoulders clip tangentially — a glancing touch, small stumble, NOT a hit; the @image2 cup jolts free; the @image3 sunglasses lift off the head and tumble; @image1 keeps the phone in the left hand. 6-11s BRAND HERO 1, HYPER ZOOM + ULTRA SLOW MOTION: rush into an extreme macro of the airborne @image2 cup nearly frozen in the air, its brand NAME filling the entire frame, razor-sharp and perfectly readable; a hard highlight sweeps slowly across the lettering so each letter GLEAMS, coffee droplets suspended and glinting. Hold long on the readable name. 11-16s BRAND HERO 2, HYPER ZOOM + ULTRA SLOW MOTION: match-cut to an extreme macro of the @image3 sunglasses turning almost imperceptibly, the brand logo filling the frame, razor-sharp; a highlight sweeps slowly across the logo so the name GLEAMS, lenses flaring. Hold long. 16-18s speed eases back: wider, @image1's smile turns to wide eyes and parted lips, phone still in the left hand; the second person recoils with an apologetic open palm reaching to steady them AFTER contact; the dog takes one startled step. Spontaneous surprise, not anger or fear. 18-20s FINAL: the @image2 cup and @image3 sunglasses settle together on the pavement in the spilled coffee, a slow rack focus drifting between the readable @image2 brand name and the @image3 brand name — both crisply accented — then hold and end. CAMERA: gimbal backward track and tilt up to open, handheld through the bump, two deliberate hyper-zoom pushes locking onto the brand names in ultra slow motion, ease wide for reactions, finish on a low static two-brand shot on the ground; natural autofocus, realistic exposure. No impossible camera move. Realistic physics on the stumble, the liquid arc, the falling sunglasses and the dog; natural motion blur, cinematic slow-motion, no soap-opera look. Exactly two hands each. SOUND: real street ambience — footsteps, distant traffic, faint chatter, dog tags jingling, a soft gasp on contact, a deep whoosh as the ultra-slow-motion brand shots engage, a crisp settle as the cup and glasses land; only @image1's single line "Hello?" spoken into the phone; no other dialogue, no invented language. NEGATIVE: the phone tucked under the arm or armpit, the phone not held in the hand at the ear, @image1 talking to themselves with no phone, or mumbling to no one; any aggression — lunge, squared shoulder, shove, body-check or punch; pre-arranged anger or fear; @image1's identity or face changing; the second person resembling or duplicating @image1; a third hand, extra arm, more than two hands, extra or missing fingers; the coffee cup or sunglasses differing from @image2 and @image3, changing brand, shape or logo, garbled, blurry, warped or misspelled brand names or logos, or invented extra branding, storefront ads or watermark; on-screen text, captions, subtitles or numbers other than the real logos on @image2 and @image3; the coffee spilling unrealistically; the cup, sunglasses or dog vanishing, dissolving, fading or teleporting — every exit is physical; plastic or beauty-filtered skin; a scene cutting off mid-action or jumping to an unrelated scene.

1 step
Seedance 2.520s · 9:16

How it works

How to make an AI video effect

  1. 1 Pick a workflow

    Browse the workflows. Every one shows the finished video first.

  2. 2 Add your photo

    Upload one clear photo. The workflow keeps your face in every shot.

  3. 3 Get your video

    The workflow runs its own prompt. Your video lands in your library.

Prompt anatomy

Inside an AI video effect prompt

Latte Bullet-Time

The setting: A crisp autumn afternoon on a tree-lined city path takes an unexpected turn when The trigger: a passing cyclist clips your avatar's shoulder mid-stride. In that instant, The effect: time suspends — What freezes: earbuds, a latte cup and a small parcel are caught in extreme slow-motion bullet-time, rotating through warm autumn light as coffee arcs and leaves drift mid-air. The release: Time resumes, everything falls, and Your face: your avatar turns to camera with a perfectly deadpan "did that really just happen" expression. A polished, Length: 15-second cinematic bullet-time moment blending high-fashion editorial style with a touch of everyday comedy.

Remix it

  • Swap the latte and the parcel for your own product.
  • Change "bullet-time" to "time-stop" for a full freeze.
  • Keep "your avatar turns to camera". That line casts you.
Open this workflow

What people make with AI video effects

  • Post a moment that stops the scrollFreeze time mid-step. Let it all fall on the beat.
  • Step off a billboardYou are the ad. Then you throw the product.
  • Turn a walk into a sceneSlow-mo, a morph or a glance that stops the park.
  • Open a launch with a revealA drop, a burst, then the product in frame.

AI video effects, with you in the shot

These workflows build an effect around one photo of you. You stop time with a glance on a park path. A cyclist clips your shoulder and the latte hangs in bullet-time. A shoe kick cuts a bedroom into a night street. A storm painting turns into a real sea and carries you through it. Each effect is directed in the prompt, beat by beat, so it lands the same way every time you run it.

AI VFX without an editing timeline

Most AI VFX tools filter a clip you already filmed. Here the effect is part of the scene from the first frame. You upload a photo, the workflow renders the set, the camera move and the effect together, and the result is one video. Nothing to key, track or composite.

Visual effects for products

Some effects are built for a product instead of a face. You pose on a billboard, climb out of it and throw the product into the street. A two-tone sphere rolls, bursts and becomes your car. Upload a clean product photo and the workflow keeps its shape and label through the effect.

Read the prompt, change the effect

Every workflow shows its full prompt. Change the setting, the object that freezes or the moment time resumes, and run it again. Your photo stays the reference every time.

Questions

Frequently asked questions

Moments a camera cannot film: time stopping mid-step, bullet-time around a dropped latte, a T-rex that turns into a macaw, a billboard model who climbs out and throws the product. Each workflow on this page writes one effect into its prompt and renders it around the photo you upload, so you are inside the effect.

Pick an effect workflow and upload one clear photo of your face. The workflow already holds the scene, the camera and the effect beat. It renders a new video with you in it, rather than filtering a clip you filmed. Change a line of the prompt to move the effect somewhere else.

Yes, on the workflows built for it. Product drops, reveals and morphs ask for a product photo and keep its shape, colour and label the same in every shot. A workflow shows which photos it needs before you run it, so you know what to upload first.

Yes. Every workflow shows its full prompt: the setting, the trigger, the effect and the timing. Read it, copy the parts you like, or run it as it is with your photo. That makes this page a library of AI video effects that already work.

Pick a workflow. Add your photo.

Every workflow is a finished scene. You bring the face.

Pick a workflow