Wan 3.0 prompt library

Cinematic

Rain on a night platform

A locked-off wide shot, then a slow push, with train brakes in the audio.

Cinematic 16:9, 8 seconds, realistic live action, no titles, no logos, no readable signs.
A nearly empty elevated train platform at night. Wet concrete, a yellow safety line, sodium lamps, and a fine rain that reads as individual drops in the foreground. One adult in a charcoal coat waits with a closed umbrella.
0–3s: locked wide shot. The person shifts their weight. Rain ticks on the umbrella canopy.
3–6s: very slow push-in to a medium shot. A distant train light blooms through the rain. They glance down the track, then back to the platform edge.
6–8s: hold on their face in profile as the train wind lifts the coat hem. They do not board.
Audio: rain on concrete, a low electrical hum, and a far brake squeal. No speech. No music.

Cinematic

Kitchen, one window

Hard morning sidelight, a kettle, and a quiet hand-held drift.

Cinematic 4:3, 8 seconds, realistic, natural light only, no titles, no logos.
A small kitchen just after sunrise. One window is the key light. Dust is visible in the beam. A steel kettle sits on a gas burner. An adult in a plain sweater stands at the counter with their back mostly to camera.
0–2s: wide shot from the doorway. The burner ticks and the flame catches.
2–5s: slow handheld drift to the window side. Steam starts. Their hand turns the kettle handle a few degrees.
5–8s: close-up on the spout and the hand, then a small tilt up to the side of their face in the window light. They breathe out. No smile for the camera.
Audio: gas tick, a soft boil, a mug set down off screen, room tone. No music, no speech.

Cinematic

Heat over a two-lane road

Telephoto heat shimmer and a car that never quite arrives.

Cinematic 16:9, 10 seconds, anamorphic feel, realistic, no titles, no logos, no readable license plate.
A straight two-lane desert road at late afternoon. Orange dust, a faded center line, telephone poles shrinking to the horizon. Heat shimmer sits on the asphalt. A pale sedan is far away, driving toward camera.
0–4s: locked telephoto wide. The car grows slowly. A dust devil crosses the shoulder and dies.
4–7s: the camera pans a few degrees with the car, still distant. A truck in the opposite lane passes and leaves a ribbon of dust that the sedan drives through.
7–10s: the sedan remains mid-distance. The shot ends before it reaches the lens. Hold on the empty foreground road.
Audio: wind across the mic, a thin tire hiss, one distant horn that is not a melody. No music, no speech.

Cinematic

Pool lights after close

Underwater reflections on a ceiling, one swimmer, no splash gag.

Cinematic 16:9, 8 seconds, realistic, cool practical light, no titles, no logos.
An indoor pool after hours. Lane ropes are stacked. Underwater lights turn the ceiling into moving caustics. One adult swims a slow breaststroke in the near lane, away from camera, then turns at the wall.
0–3s: high angle from the deck. Only the swimmer and the empty water.
3–6s: lower angle at water level. The turn is unhurried. A breath, then they push off.
6–8s: the camera stays on the ripple after they leave the frame. Caustics keep moving.
Audio: a single breath, water slap, the hum of the pool filter. No music, no whistle, no speech.

Cinematic

Stairwell, one fluorescent tube

A vertical 9:16 climb with a flicker and a metal handrail.

Cinematic vertical 9:16, 8 seconds, realistic, no titles, no logos, no readable posters.
A concrete stairwell lit by one flickering fluorescent tube. Pale green walls, a steel handrail, scuffed steps. An adult in work boots climbs toward camera, which is one landing above them.
0–3s: they enter from the bottom of frame, hand on the rail, head down.
3–6s: a closer view as they reach the landing. The tube flickers twice. They pause, look up toward the camera height, not into the lens, and keep climbing.
6–8s: they pass out of the top of frame. The camera holds on the empty landing and the buzzing light.
Audio: boots on concrete, the rail creak, fluorescent buzz. No speech, no alarm.

Cinematic

Greenhouse at dusk

Condensation, a hose, and a slow lateral move through leaves.

Cinematic 16:9, 8 seconds, realistic, no titles, no logos, no readable plant tags.
A glass greenhouse at dusk. Condensation on the panes, warm work lights inside, blue air outside. Tomato vines and a coiled green hose. An adult in a canvas apron checks one vine, then crouches.
0–3s: slow lateral move behind the leaves, the person half hidden.
3–6s: they come clear, pinch a yellow leaf, and drop it into a bucket.
6–8s: push through a gap in the vines to a close shot of wet glass as a drop runs down. The person is soft in the background, still working.
Audio: a hose drip, glass tick, cloth movement. A single far bird. No music, no speech.

Ads and product

Glass bottle pour

A 6-second product beat: condensation, pour, and a hero hold.

Square 1:1 product shot, 6 seconds, studio realistic, no titles, no logos, no label text.
A clear glass bottle of still water on a pale stone surface. Soft north-light look, one gentle shadow. Condensation beads are already on the glass. A short tumbler sits to the right.
0–2s: macro on the shoulder of the bottle. A bead runs.
2–4.5s: a hand with no jewelry tilts the bottle and pours a clean stream into the tumbler. No splash onto the lens.
4.5–6s: the bottle returns upright. Hold on bottle and glass, both sharp. The hand leaves frame.
Audio: glass touch, a short pour, one last drip. No jingle, no voiceover.

Ads and product

Sneaker on wet asphalt

Vertical product turntable energy without a plastic turntable.

Vertical 9:16, 6 seconds, realistic product film, no titles, no logos, no brand marks, no readable text on the shoe.
One plain white leather sneaker on wet dark asphalt at night. A single hard light from the left makes a clean highlight along the sole. The background is out of focus city bokeh with no signs.
0–2s: low angle, the shoe is still. A drop falls from above and hits beside it.
2–4.5s: the shoe rotates slowly a quarter turn as if nudged, showing the side and then the heel. It does not hop or dance.
4.5–6s: hold on the heel highlight. The wet ground reflects the shoe.
Audio: one water drop, a soft rubber scrape, distant traffic. No music, no voice.

Ads and product

Dropper on a bathroom shelf

A calm 8-second serum spot with a real bathroom, not a plastic set.

Vertical 9:16, 8 seconds, realistic, soft daylight, no titles, no logos, no label text.
A frosted glass dropper bottle with a blank white label sits on a stone bathroom shelf. A folded gray towel and a small plant are behind it, out of focus. No mirror selfie, no face.
0–2.5s: still life. The camera creeps in.
2.5–5.5s: a hand unscrews the dropper, lifts it, and lets one drop fall back into the bottle. The drop is sharp.
5.5–8s: the dropper returns. Hold on the bottle, three-quarter view.
Audio: a cap thread, a single drop, room tone. No music, no voiceover.

Ads and product

Handmade mug, steam

A tabletop ad that sells the object by how it sits in the hand.

4:3, 8 seconds, realistic tabletop, morning window light, no titles, no logos, no text on the mug.
A speckled ceramic mug, off-white glaze, on a wood table. Steam rises. A linen napkin is folded once. No laptop, no phone.
0–3s: wide enough to see the table corner. A hand enters and wraps the handle.
3–6s: they lift the mug a few centimeters, turn it so the glaze catches the window, and set it down without a clink harder than a tap.
6–8s: close-up of the rim and steam. The hand rests beside the mug, not covering it.
Audio: ceramic on wood, a quiet sip implied but not shown, room tone. No music, no speech.

Ads and product

Desk lamp, one click

A hardware spot built around the click and the pool of light.

16:9, 5 seconds, realistic, no titles, no logos, no brand marks.
A matte black desk lamp on a dark oak desk in a dim room. A closed notebook with a blank cover lies beside it. The lamp is off.
0–1.5s: wide shot of the desk. A hand enters.
1.5–3s: the finger finds the switch. Click. A warm pool of light hits the notebook and the wood grain. The rest of the room stays dim.
3–5s: slow push along the arm of the lamp to the lit page. The hand leaves.
Audio: the click, a tiny spring, then near silence. No music, no voice.

Reference to video

Two cooks, one bench

A text brief that keeps two people distinct. The link does not upload photos.

This brief is written so a reference photo could lock identity. The button fills the text prompt only. It does not upload a photo. For a locked reference-photo template, use the templates index.

Text-to-video. 16:9, 8 seconds, realistic, no titles, no logos.
If reference photos were attached, Image1 would be the person on the left and Image2 the person on the right, and neither face would be copied onto the other. This link does not attach photos, so invent two distinct adults and keep each face stable for the whole clip.
A bakery bench at opening. Flour, a wood peel, trays of plain rolls with no lettering.
0–3s: wide shot. The left person scores a roll. The right person slides a tray.
3–6s: they pass a towel between them without looking at the lens. Hands do not merge.
6–8s: both look down at the bench. Hold. Same two faces as the opening.
Audio: tray metal, a knife tap, room tone. No speech, no shop bell jingle.

Reference to video

Red coat, same person across the street

Identity held across a cutless street crossing.

This brief is written so a reference photo could lock identity. The button fills the text prompt only. It does not upload a photo. For a locked reference-photo template, use the templates index.

Text-to-video, vertical 9:16, 8 seconds, realistic, no titles, no logos, no readable signs.
Written in a reference style: one person only, the same face, hair, and red wool coat from the first frame to the last. Do not invent a second face. This link does not upload a photo.
A city crosswalk at dusk, wet paint, traffic lights as color only.
0–3s: medium shot on the curb. They button the coat.
3–6s: they step off and cross. The camera tracks backward. The face stays recognizable when a car passes behind them.
6–8s: they reach the far curb and stop. Same coat, same face. They look along the street, not into the lens.
Audio: footsteps, a walk signal beep that is a plain tone, city wash. No speech.

Reference to video

Same hands on the wheel

A reference-style brief for hands and apron, not a face swap.

This brief is written so a reference photo could lock identity. The button fills the text prompt only. It does not upload a photo. For a locked reference-photo template, use the templates index.

Text-to-video, 16:9, 8 seconds, realistic, no titles, no logos.
Reference-style constraint with no photo attached: one adult, the same hands, the same clay-stained apron, the same short hair. Do not change their face between shots because there are no cuts.
A pottery studio. A kick wheel, a lump of gray clay, a window.
0–3s: close on the hands centering the clay. The face is soft in the background and already matches the person we will see.
3–6s: tilt up to a medium shot. Same person. They lean in. The wheel slows.
6–8s: back to the hands. The clay is a simple bowl. Same apron straps.
Audio: wheel hum, a wet hand, room tone. No speech, no radio.

Reference to video

One dog, one person, hallway

Keeps a pet and a person consistent without a photo upload.

This brief is written so a reference photo could lock identity. The button fills the text prompt only. It does not upload a photo. For a locked reference-photo template, use the templates index.

Text-to-video, 16:9, 6 seconds, realistic, no titles, no logos, no collar tags with text.
Reference-style, no photo attached. One adult and one medium tan dog. The dog's coat, ears, and the person's glasses stay the same the entire time. Do not duplicate the dog.
A narrow apartment hallway with a wood floor and a single lamp.
0–2s: the dog sits. The person kneels and offers an empty hand.
2–4.5s: the dog stands and walks two steps into the hand. Tail moves once.
4.5–6s: both hold. The person's other hand stays on their knee. Same dog, same face.
Audio: claws on wood, a quiet breath, no speech, no music.

Reference to video

Fruit stall, three distinct people

A crowd of three, each with a stable face, written as text.

This brief is written so a reference photo could lock identity. The button fills the text prompt only. It does not upload a photo. For a locked reference-photo template, use the templates index.

Text-to-video, 16:9, 10 seconds, realistic, no titles, no logos, no readable price cards.
Reference-style brief with no photos attached. Exactly three adults: a vendor in a green apron behind the crates, a customer in a navy jacket in front, and a passerby in a gray cap who only crosses the background. Keep those three faces distinct. Do not copy one face onto another.
Oranges and green apples in wood crates. Morning street, soft sun.
0–3s: wide. The vendor shifts a crate. The customer points at the apples, not at a sign.
3–7s: medium shot of the exchange. The vendor nods and sets two apples into a plain paper bag with no print.
7–10s: the passerby crosses behind them and is gone. The customer takes the bag. All three faces, if still visible, match the opening.
Audio: crate wood, paper rustle, street wash. No speech, no market jingle.

30-second long take

Harbor walk, one take

Thirty seconds beside the water, no cut, weather in the audio.

One continuous 30-second take, 16:9, realistic, no cuts, no titles, no logos, no readable boat names.
Camera: a gimbal at chest height, walking beside one adult in a navy peacoat. The path is a harbor boardwalk. Fishing boats are tied up. Gulls are small in the sky. Late afternoon, wind.
0–6s: start on their boots and the wet planks, then rise to a three-quarter view from their left. They walk at a steady pace. The camera matches them. Rigging clinks.
6–12s: they pass a stack of lobster pots. A loose rope taps a post. They glance at the water, not at the lens. The horizon stays level.
12–18s: a gap between boats shows open water. Wind presses the coat. They put one hand in a pocket. The camera drifts half a step behind, then catches up. No whip pan.
18–24s: a cyclist passes on the far side of the path, out of focus, and is gone. Our person does not react much. A bell buoy sounds once, far off.
24–30s: they slow near a railing, stop, and look out. The camera stops with them and eases into a medium close-up. Hold on the water and the side of their face until the last second.
Audio: footsteps on wood, halyard clinks, wind on the mic, one buoy bell, water against pilings. No music. No speech.

30-second long take

Library aisle, one take

A quiet 30-second move from the doors to a single book.

One continuous 30-second take, 16:9, realistic, no cuts, no titles, no readable book titles, no logos.
A public reading room with high windows and dust in the light. Almost empty. One adult with a canvas tote walks from the entrance toward a side aisle. The camera follows from behind, then slides to their side. Shoes are soft-soled.
0–7s: wide, from the doors. They cross a patch of sun. A chair creaks somewhere off screen. They do not look back.
7–14s: into the aisle. Shelves on both sides. The camera is now at their shoulder. Their fingers trail the shelf without pulling a book. Footsteps get quieter.
14–21s: they stop. The camera keeps moving a short beat, then settles. They tilt their head, reading spines that must stay illegible, just color and wear.
21–27s: they draw one plain clothbound book, no foil stamping, and open it with one hand. A page turns.
27–30s: close on the open page, which is blank cream paper, and their thumb. Hold. The room tone stays.
Audio: distant HVAC, a soft shoe, paper, no voice, no music, no phone.

30-second long take

Night bus, one take

Interior of a moving bus for a full 30 seconds, window light only.

One continuous 30-second take, 16:9, realistic, no cuts, no titles, no logos, no readable ads.
Inside a nearly empty city bus at night. Fluorescent lights are off. The only light is streetlamps sliding across the windows. One adult sits by the window in the mid row, coat collar up. The camera is across the aisle, handheld but calm.
0–8s: wide enough to see empty seats. The bus sways. A handrail ticks. Their reflection is faint in the glass.
8–16s: slow push toward the window seat. Outside, a laundromat glow passes, then dark shops. They blink and look at their hands. A paper transfer is blank and stays blank.
16–23s: the bus brakes gently. They steady a canvas bag on their lap. No one else boards. The doors do not open in frame.
23–30s: they lean their head toward the glass, not asleep, just resting, and the streetlights stripe their face. Hold as the bus rolls again.
Audio: engine drone, a rail tick, a soft brake, no announcement, no music, no speech.

30-second long take

Bench saw, one take

Thirty seconds of a person measuring and marking, not a stunt with the blade.

One continuous 30-second take, 16:9, realistic, no cuts, no titles, no logos, no brand names on tools.
A small woodshop in daylight from a side window. An adult in a plain shirt stands at a bench. A hand saw hangs on the wall and is not used. They are marking a pine board with a pencil. The camera is on a short slider at bench height.
0–8s: wide. They set a steel square on the board, check it, and pick up the pencil. Sawdust is already on the floor.
8–16s: the slider moves left to right as they draw one straight line. The pencil sound is clear. They blow the dust away once.
16–23s: they step back, look down the line, and rotate the board a quarter turn. The camera eases closer. Their face is concentrated, not performing.
23–30s: close on the pencil line and the grain. Their hand rests flat on the wood. Hold. The saw stays on the wall.
Audio: pencil on pine, a board scrape, a far street sound through the window. No blade, no music, no speech.

Dialogue with audio

Two lines at a kiosk

A six-second exchange, mouths in time, no extra chatter.

16:9, 6 seconds, realistic, no titles, no logos, no readable menu.
A small outdoor coffee kiosk. Two adults: the server behind the counter in a plain apron, the customer in a gray jacket. Medium two-shot.
The server slides a lidded paper cup with no print across the counter and says, clearly, in a calm voice: "Careful, it's hot."
The customer takes the cup and answers: "Thanks. I'll walk slow."
Mouths match those words and then close. No other speech. The cup does not show a brand.
Audio: those two lines, a cup slide, street wash underneath. No music.

Dialogue with audio

One sentence over a bicycle

A repair stand, a short line, and the click of a gear.

16:9, 8 seconds, realistic, no titles, no logos, no brand marks on the bike.
A narrow repair shop. One adult stands at a work stand with a plain black bicycle, rear wheel toward camera. They spin the pedal once, listen, and then look toward someone just out of frame.
They say, even and unhurried: "The click is gone. You can take it."
After the line they wipe their hands on a cloth and nod. No second voice answers in this clip. Mouth movement matches that sentence only.
Audio: the sentence, a freewheel tick that stops, a cloth brush. Workshop room tone. No music.

Dialogue with audio

Elevator, a short apology

Vertical frame, two people, one overlap of a word.

Vertical 9:16, 8 seconds, realistic, no titles, no logos, no floor numbers that can be read.
An elevator with brushed metal doors, already open a crack at the start. Two adults inside: one holding a cardboard box with no print, one with empty hands.
0–3s: the person with empty hands reaches for the door close, then stops as the other shifts the box.
They say: "Sorry — your floor?"
The person with the box answers: "Next one. I'm fine."
3–8s: a small nod from both. Doors stay as they were. No joke, no extra lines. Mouths match the words.
Audio: the two lines, a soft box corner tap, elevator hum. No chime melody, no music.

Dialogue with audio

Loading dock, end of shift

A practical line about a door, with a pallet jack in frame.

16:9, 10 seconds, realistic night exterior, no titles, no logos, no readable labels on boxes.
A loading dock under one sodium lamp. Two adults in plain work jackets. A manual pallet jack holds three blank cartons. Breath shows in the cold.
The person nearer the door checks a padlock and says: "Leave it. The morning crew has the key."
The other person answers: "Then I'm done. See you Thursday."
They do not shake hands. One pulls the pallet jack back a half step so it is square. Mouths match only those lines.
Audio: the dialogue, a pallet wheel, a jacket zip, distant road. No music.

Dialogue with audio

Window seat, one question

Quiet speech against cabin wash, no airline branding.

16:9, 8 seconds, realistic, no titles, no logos, no airline marks, no readable safety card.
A dim cabin. Two adults in the window row, plain sweaters, seat-back screens off and blank. Night outside the oval window. The camera is from the aisle, a calm medium shot.
The aisle person leans slightly and asks, quietly: "Do you want the shade down?"
The window person looks outside, then answers: "Leave it. I like the lights."
They both face forward again. No more words. Mouths match the lines. The shade stays up.
Audio: those two lines close-miked but still soft, a steady cabin rush underneath, no chime, no music.