Wan 3.0 prompt templates

Wan 3.0 Prompts: Copy-Ready Video Templates

Browse practical Wan 3.0 prompts for ads, UGC, product films, multi-shot stories, and audio-led scenes. Copy a complete brief, replace the variables, and open it in the generator.

15-template libraryUpdated September 12, 202615 copy-ready prompts

Start here

How to turn a template into your prompt

Choose a template by production job, not by visual fashion. An ad prompt should make one benefit or mechanism easy to see. A UGC prompt needs a believable performance and camera relationship. A product prompt protects geometry and materials. A multi-shot prompt manages sequence and continuity. An audio prompt assigns speech, effects, ambience, and music to specific moments. Starting with the right job removes more ambiguity than adding a long style list.

Replace bracketed variables with concrete nouns and observable actions. Instead of “[PRODUCT] looks premium,” describe the shape, material, moving part, and light behavior that communicate its value. Instead of “[CREATOR] is excited,” describe the pause, glance, breath, or gesture that makes excitement visible. Read the finished prompt once as a shot list: every subject, prop, movement, camera instruction, and sound cue should be physically compatible.

Open a template in the wan-3.ai generator, then confirm input mode, model, duration, aspect ratio, resolution, audio, and references before submission. Text-to-video prompts can define the whole scene. Image-to-video templates assume you will upload a source frame and should focus on movement plus preservation. Reference-led templates use numbered image, video, or audio tokens only after matching files are uploaded.

  • Replace every bracketed variable before generating.
  • Delete directions that compete with the one message or action.
  • Choose aspect ratio for the destination, not after the composition is made.
  • Save the prompt and settings with every version worth reviewing.

Control

Protect the details your reviewer will reject if they drift

A reusable prompt separates creative freedom from fixed requirements. Put protected traits in one sentence: product silhouette, number of parts, label position, wardrobe, face shape, location layout, or the opening and closing frames. Then describe where the model may explore—background atmosphere, secondary motion, light progression, or a small camera adjustment. This boundary is especially important when animating an approved image.

Avoid asking generated footage to carry precise legal copy, prices, statistics, interface text, or packaging language unless the workflow has been specifically tested for that requirement. Leave intentional negative space and add exact typography during editing. A prompt can direct the composition for a headline without pretending the generated frame is the safest place to produce final campaign wording.

Treat people, brands, voices, music, and uploaded assets as production inputs that need permission. The templates below use fictional, unbranded scenarios. When adapting them, verify that your team has the rights required for likenesses, source footage, logos, music cues, and commercial distribution. Model capability does not replace that review.

Iteration

Run a controlled test instead of random variations

For a first pass, choose the shortest duration that can clearly contain the action. Review whether the subject, message, motion, camera, sound, and end state work before increasing visual complexity or output settings. When a result is close, change the clause that controls the largest problem. Clarify motion speed if an action is weak; simplify camera movement if framing drifts; reduce pose or viewpoint change if recognition is unstable.

Keep a compact version note: original template, customized prompt, input mode, references, model, settings, result, and one variable changed. A useful A/B test might compare locked camera with a slow push-in while everything else remains fixed. Comparing two entirely different prompts may produce two attractive clips, but it will not teach the team which decision improved the result.

  • First pass: validate the message and main action.
  • Second pass: fix the largest continuity or camera problem.
  • Final pass: confirm sound, end-frame usability, format, and rights.
  • Do not interpret a successful example as a guaranteed repeatable output.

Prompt library

Copy a complete brief, then make it yours

Replace bracketed details, remove directions that do not serve the shot, and confirm the selected mode before generation.

Ads

Ad prompts built around one visible benefit

Use motion to demonstrate the promise, reserve clean space for final copy, and land on an editable frame.

01

Ads

Problem-to-solution social ad

A clear two-state story for a service or household product without relying on generated claims.

Create a 12-second 9:16 advertisement for [PRODUCT OR SERVICE BENEFIT], with no visible brand name or generated text. Shot 1 [0–5s]: [TARGET USER] struggles with [SPECIFIC, VISIBLE PROBLEM] in [REALISTIC LOCATION]; handheld medium framing, slightly cool natural light, only practical room sounds. Shot 2 [5–10s]: hard cut as [PRODUCT OR SERVICE ACTION] resolves the same problem in one readable motion; stabilize the camera and warm the light subtly. Shot 3 [10–12s]: clean close-up of [RESULT OR PRODUCT] with generous empty space in the upper third for copy added later. Keep the person, wardrobe, props, room layout, and direction of light consistent. Audio: one clear transition sound at the solution moment, then a restrained upbeat music cue. No subtitles, prices, statistics, logos, extra hands, or unrelated background action.

Opens with this prompt ready to editUse this prompt on wan-3.ai
02

Ads

Feature proof in one continuous shot

Show one mechanism from start to finish so the viewer understands the benefit without a voiceover.

Single continuous demonstration of [PRODUCT] performing [ONE FEATURE]. Begin with [PRODUCT] centered on [SURFACE] in a clean real-world setting. [HAND OR USER] enters from frame right, performs [EXACT OPERATION] once, and pauses so the completed state is easy to inspect. Use a chest-height three-quarter camera and a slow 15-degree lateral slide that never hides the mechanism. Preserve [PRODUCT SHAPE], [MATERIAL], [COLOR], [CONTROL POSITION], and the number of parts. Lighting should reveal edges and texture without dramatic color changes. Sound: natural handling, one precise mechanical response, low room tone; no voice and no music. No duplicate product, extra controls, visible text, logo, smoke, sparks, or impossible movement. End on the completed state with the hand out of frame and hold for one second.

Opens with this prompt ready to editUse this prompt on wan-3.ai
03

Ads

Seasonal mood ad with a product anchor

Build atmosphere around a consistent hero object while keeping the product visible in every beat.

Create a 15-second widescreen seasonal brand film for an unbranded [PRODUCT]. Overall mood: [SEASON OR OCCASION], tactile, restrained, human. Shot 1 [0–5s]: macro view of [SEASONAL MATERIAL OR ACTION] revealing the edge of the product; slow push-in. Shot 2 [5–10s]: medium shot of [PERSON] using the same product in [LOCATION], with one simple gesture that communicates [BENEFIT]; gentle side track. Shot 3 [10–15s]: product returns to a centered still life surrounded by only two relevant seasonal elements; fixed camera. Preserve product dimensions, color, surface finish, and packaging across shots. Use [PALETTE], believable practical light, and shallow focus only where it does not hide the feature. Audio moves from tactile close sounds to subtle room ambience; a quiet [INSTRUMENT] theme enters after Shot 1. No spoken claim, text, logo, duplicate object, confetti, or exaggerated particles. Hold the final composition for two seconds.

Opens with this prompt ready to editUse this prompt on wan-3.ai

UGC

UGC prompts with believable performance beats

Direct one person, one situation, and one clean line instead of simulating an over-edited influencer montage.

04

UGC

Hook, demo, reaction

A three-beat creator format that keeps the product and room consistent across cuts.

Create a 15-second vertical UGC video in [ROOM]. Shot 1 [0–4s]: medium close-up, natural phone framing; [CREATOR DESCRIPTION] looks at camera and says, “[SHORT PROBLEM-LED HOOK],” with a conversational pace and no exaggerated expression. Shot 2 [4–11s]: hard cut to an over-shoulder close-up as the creator uses [PRODUCT] to complete [ONE VISIBLE TASK]; show the full action without time-lapse or hidden steps. Shot 3 [11–15s]: return to the original framing; the creator glances at the finished result, then says, “[BRIEF, NON-NUMERIC REACTION].” Preserve face, hands, clothing, product color, room layout, and daylight direction. Audio: clean speech, natural handling sounds, low room tone; no music under dialogue. No captions, logo, beauty filter, extra people, duplicate product, or sudden lens changes. End on a small natural nod rather than a posed freeze.

Opens with this prompt ready to editUse this prompt on wan-3.ai
05

UGC

Day-in-the-life insert

Place the product inside a real routine and keep the action small enough to remain credible.

Vertical day-in-the-life clip of [CREATOR] using [PRODUCT] during [SPECIFIC ROUTINE]. Begin with a phone-height wide shot as the creator enters [LOCATION] carrying [RELEVANT PROP]. Cut to a close-up of one hand placing or activating the product, then return to a medium shot as the creator continues the routine without addressing the camera. Natural available light, slight handheld drift, realistic background that remains quiet and uncluttered. Keep the creator’s hairstyle, clothing, product shape, and prop positions consistent. Sound: footsteps, fabric, product handling, and location ambience; no dialogue, no music for the first half, then a low-key rhythmic track fades in. Do not show app screens, generated labels, brand logos, extra fingers, duplicate objects, or impossible product behavior. End with the product still visible but not held toward the lens.

Opens with this prompt ready to editUse this prompt on wan-3.ai
06

UGC

Simple myth check

A direct-to-camera setup followed by one visual test and a measured conclusion.

Create a 14-second vertical creator video testing the statement “[COMMON ASSUMPTION]” about [PRODUCT CATEGORY]. Shot 1 [0–4s]: waist-up phone framing in [LOCATION]; [CREATOR] says, “I keep hearing that [SHORT ASSUMPTION], so I tried one simple test.” Shot 2 [4–10s]: fixed close-up of the creator performing [SAFE, VISIBLE TEST] once, with the result fully in frame and no jump that hides the process. Shot 3 [10–14s]: return to the original framing; the creator says, “[CAUTIOUS OBSERVATION WITHOUT A UNIVERSAL CLAIM].” Keep product, clothing, hands, background, and light consistent. Clean dialogue, natural test sounds, no background music, captions, statistics, logo, dramatic reaction, or extra participants. Do not imply laboratory validation. Finish on the result beside the creator, with neutral expression and room for an editor-added note.

Opens with this prompt ready to editUse this prompt on wan-3.ai

Product

Product prompts that protect shape, surface, and count

Lock recognizable geometry before asking for camera movement, environmental effects, or transformation.

07

Product

Controlled hero orbit

A slow, bounded orbit designed to reveal material without inventing a new back side.

Single studio hero shot of one unbranded [PRODUCT] on [SURFACE]. The product remains motionless while the camera performs a slow clockwise orbit of approximately 30 degrees, moving from front three-quarter to side three-quarter view; keep the camera height and focal distance constant. Preserve the exact silhouette, [KEY COMPONENTS], [MATERIAL], [COLOR], seams, control placement, and object count. A narrow key light reveals [TEXTURE], with soft fill and one controlled reflection that travels across the front edge. Background is a smooth [COLOR] gradient with no horizon clutter and generous negative space on [SIDE]. Sound: quiet studio tone and one subtle material-specific effect; no voice or music. No floating particles, liquid, smoke, text, logo, hand, duplicated part, deformation, or explosive transition. End with the product fully visible and sharp.

Opens with this prompt ready to editUse this prompt on wan-3.ai
08

Product

Ordered exploded assembly

Best for a conceptual component reveal when the exact assembly order and object count are known.

Create a clean conceptual assembly film for [PRODUCT] using exactly [NUMBER] visible components: [LIST COMPONENTS IN ORDER]. Begin with the assembled product centered against a neutral background. The parts separate slowly along one horizontal axis while retaining their orientation and spacing evenly; pause in an exploded view, then return in reverse order and lock into the original silhouette. Fixed camera, orthographic-feeling three-quarter view, no rotation, no hidden component, no new fastener, and no change to scale. Use soft technical lighting with crisp edges and restrained [ACCENT COLOR] highlights. Audio: one low mechanical separation sound, quiet air movement, and precise clicks as each part reconnects; no voice or music. Do not generate labels, diagrams, hands, sparks, extra pieces, or transparent internals. Hold the fully reassembled product for the final second.

Opens with this prompt ready to editUse this prompt on wan-3.ai
09

Product

Image-to-video light pass

Upload a final product still and animate only light, reflection, and one small environmental detail.

Animate the uploaded product still while preserving the exact opening composition. Keep product silhouette, proportions, material, color, packaging, label shapes, object count, camera angle, crop, surface, and background fixed. The product does not rotate or move. Add a soft narrow highlight that travels once across [MATERIAL SURFACE] from left to right, a subtle reflection shift on the nearest edge, and [ONE SMALL ENVIRONMENTAL MOTION] in the background. Locked camera with no push, pull, pan, orbit, shake, focus breathing, or reframing. Audio: minimal location ambience and one quiet material sound; no dialogue or music. No new text, logo, prop, hand, liquid, smoke, duplicate object, or geometry change. End with the same composition and a slightly calmer light state, holding steady for one second.

Opens with this prompt ready to editUse this prompt on wan-3.ai

Multi-shot

Multi-shot prompts with continuity written in

Use time ranges, transition logic, and a short continuity list so the sequence feels like one production.

10

Multi-shot

Arrival, discovery, departure

A location story with one traveler, one prop, and a consistent direction of movement.

Overall direction: a quiet 21-second travel story about [TRAVELER DESCRIPTION] discovering [LOCATION DETAIL], moving consistently from left to right. Shot 1 [0–7s]: wide establishing frame as the traveler enters [LOCATION] carrying [PROP]; slow lateral track, overcast morning ambience. Shot 2 [7–14s]: hard cut to a medium close-up from the same screen direction as the traveler notices [DISCOVERY], stops, and reaches toward it without touching; camera makes a very small push-in. Shot 3 [14–21s]: match cut on [SHARED SHAPE OR MOTION] to a rear wide shot as the traveler continues toward [DESTINATION]; fixed camera and a final glance back. Preserve face, [WARDROBE], prop, weather, ground condition, and color palette. Audio: footsteps and location ambience throughout; one restrained [INSTRUMENT] phrase begins at the discovery. No dialogue, captions, landmark signage, crowd close-ups, time-of-day change, or teleporting objects. End with the traveler small in the frame.

Opens with this prompt ready to editUse this prompt on wan-3.ai
11

Multi-shot

Food process without continuity jumps

A compact recipe sequence that keeps ingredients, vessel, and surface consistent.

Overall direction: an 18-second tactile preparation film for [DISH OR DRINK] on the same [SURFACE], with natural color and no visible brand. Shot 1 [0–6s]: overhead close-up as exactly [INGREDIENT COUNT] prepared ingredients are arranged beside [VESSEL]; locked camera and crisp handling sounds. Shot 2 [6–12s]: side macro view as [PRIMARY ACTION] changes the texture from [START STATE] to [END STATE]; slow push-in, no skipped tool swap. Shot 3 [12–18s]: three-quarter medium shot as the finished item is placed in the same vessel and one final garnish is added; fixed camera, steam or condensation moves naturally. Preserve vessel shape, surface grain, ingredient colors, hand appearance, and daylight direction. Audio: cutting, pouring, stirring, or ceramic contact at realistic volume; soft room tone; no music until a single warm chord in the last three seconds. No labels, text, extra utensils, duplicated ingredients, impossible melting, or messy background. Hold the finished result.

Opens with this prompt ready to editUse this prompt on wan-3.ai
12

Multi-shot

Concept transformation with a match cut

Connect two worlds through one repeated form instead of an unexplained morphing spectacle.

Overall direction: a 16-second visual metaphor connecting [FIRST IDEA] to [SECOND IDEA] through the repeated circular shape of [OBJECT A] and [OBJECT B]. Shot 1 [0–6s]: centered close-up of Object A in [ENVIRONMENT A]; it completes [SIMPLE ACTION] while the camera remains fixed. Shot 2 [6–10s]: match cut at the exact moment Object A fills the frame, preserving position, scale, and direction as it becomes the circular form of Object B; one smooth transition only. Shot 3 [10–16s]: pull back slowly to reveal Environment B as Object B performs [RELATED ACTION]. Keep the circle centered, use a compatible [PALETTE] across both settings, and maintain the same direction of light. Audio bridges the cut with [SHARED SOUND QUALITY], then resolves into Environment B ambience. No text, logo, extra transformation, camera shake, unrelated object, or color flash. End on a balanced wide composition with Object B still recognizable.

Opens with this prompt ready to editUse this prompt on wan-3.ai

Audio

Audio prompts that specify source, timing, and restraint

Write sound as production direction: who speaks, what makes each effect, what fills the room, and when music enters.

13

Audio

One-line performance

A short line with a pre-speech action, delivery direction, and protected silence afterward.

Single continuous medium close-up of [SPEAKER DESCRIPTION] in [QUIET LOCATION]. Before speaking, the person completes [SMALL PHYSICAL ACTION], takes one visible breath, and looks toward [LISTENER OR CAMERA POSITION]. They say in [LANGUAGE AND ACCENT], “[EXACT SHORT LINE],” with [EMOTION], a measured pace, and a natural pause after [WORD OR PHRASE]. Keep mouth movement aligned with the line and do not add speech. The camera makes a slow, nearly imperceptible push-in while preserving eye level and composition. Audio layers: low [LOCATION] ambience throughout; the physical action is clearly audible before the line; no music under speech; two seconds of clean room tone after the final word. Preserve face, hair, clothing, posture, and light direction. No captions, second speaker, voiceover, echo effect, background conversation, or dramatic reaction. End on the held look.

Opens with this prompt ready to editUse this prompt on wan-3.ai
14

Audio

Ambient world without dialogue

Build atmosphere from physical sound sources while keeping music out of the scene.

Create a 14-second single-shot atmosphere study of [LOCATION] at [TIME AND WEATHER], with no people speaking. Begin on a wide static frame, then track forward slowly along [PATH] while keeping [VISUAL ANCHOR] near the center. Primary motion: [ONE ENVIRONMENTAL ACTION]. Secondary motion: [TWO SMALL RESPONSES]. Build the soundtrack only from identifiable sources: [NEAR SOUND] in the foreground, [MID-DISTANCE SOUND] across the space, and [DISTANT SOUND] beyond the frame. Let volume and stereo position change naturally as the camera moves. No music, voice, narration, synthetic riser, captions, logo, sudden weather change, or unrelated animal. Use realistic light, restrained color, stable geometry, and continuous direction of movement. End as the camera stops beside [FINAL DETAIL], allowing the ambience to continue for one second.

Opens with this prompt ready to editUse this prompt on wan-3.ai
15

Audio

Beat-synced material montage

Use three real object sounds as edit points, then introduce music only after the visual rhythm is established.

Create a 12-second material montage for [CRAFT OR PRODUCT CATEGORY]. Shot 1 [0–4s]: macro close-up of [ACTION 1]; cut on the exact sound of [IMPACT OR TOOL]. Shot 2 [4–8s]: side close-up of [ACTION 2]; cut on [SECOND DISTINCT SOUND]. Shot 3 [8–12s]: wider reveal of the finished [OBJECT] as [FINAL ACTION] creates [THIRD SOUND], then hold. Keep the same hands, material color, work surface, tool set, and warm side light across all shots. Audio begins with only clean close-recorded effects and low workshop ambience. After the second cut, add a restrained rhythm built from the three recorded sound qualities; no vocal sample, melody, or heavy bass. No dialogue, captions, logo, extra fingers, duplicate tools, missing process step, sparks, or fast camera spin. End with the finished object centered and the final sound decaying naturally.

Opens with this prompt ready to editUse this prompt on wan-3.ai

Practical answers

Questions creators ask before a run

Can I use these Wan 3.0 prompts unchanged?+

You can copy any template, but replace the bracketed variables and remove directions that do not fit your concept. A template becomes useful when its subject, action, camera, sound, duration, and final frame describe one compatible production. Review all rights and factual claims before using a generated result commercially.

Which prompt category should I start with?+

Choose the job. Ads demonstrate one benefit, UGC centers a believable creator performance, Product prompts protect shape and materials, Multi-shot prompts organize a sequence, and Audio prompts make timing and sound sources explicit. A concept can cross categories, but the first template should solve its hardest production decision.

How do I send a template to the generator?+

Select “Use this prompt on wan-3.ai.” The link opens the video generator with Wan 3.0 and the relevant input mode selected, and the complete prompt is placed in the editor. You can revise it before submitting. Uploaded images, video, and audio still need to be added in the generator.

What should I remove from a template?+

Remove any camera move, sound layer, prop, style, or constraint that does not change the intended message. Delete incompatible instructions first. If the shot is a quiet product demonstration, it may not need music, particles, a moving camera, and several locations. Fewer well-chosen decisions are easier to judge and revise.

How often will this Wan 3.0 prompt library change?+

The library is designed for regular additions and editorial updates. New templates should add a distinct production job rather than a shallow variation of an existing prompt. Product limits and generator links must be rechecked when the available models or controls change.

Keep building the workflow

Product references

Model limits and prompt structures were checked against the current wan-3.ai generator contract and the primary documentation below. Creative recommendations are editorial guidance, not a guarantee of a specific output.