Best Seedance 2.0 Prompts: 50 Copy-Paste Ideas [2026]

Seedance 2.0 is strongest when you give it a scene to direct. A weak prompt says "cinematic product video, beautiful lighting." A stronger prompt says "A matte black speaker sits on wet concrete as rain beads on the grille, slow push-in, low thunder and water hiss."

That tiny difference changes everything. The first prompt is a mood. The second prompt is a job. Seedance can stage it, light it, move the camera, and place audio against it.

Use this page as a prompt library, but do not copy blindly. Pick the category closest to your clip, then adjust the subject, camera move, aspect ratio, and duration. The point is not to memorize one magic sentence. It is to learn the shape of prompts that survive generation. The prompt structure section below explains the formula behind every example, and if you are writing from scratch or starting from a still image, the text-to-video and image-to-video sections cover the workflow and settings for each.

Update, August 2026: Seedance 2.5 is now live on Morphed with 30-second single takes, audio generated in the same pass, and reference briefs of up to 50 assets. Every prompt structure on this page still works on 2.5, and the new capabilities have their own dedicated guide: Seedance 2.5 prompts.

Before You Start

Think of Seedance as a small production crew. It needs the same things a real crew would need: what is in frame, what changes, where the camera is, how long the beat lasts, and what the viewer hears. If you only describe the final image, Seedance has to invent the motion by itself. That is when clips become floaty, over-edited, or weirdly static.

For every prompt below, you can safely swap the subject while keeping the structure. A shoe can become a perfume bottle. A cafe can become a studio. A cyclist can become a runner. The important part is the motion logic.

What Prompt Structure Works Best?

The most reliable Seedance 2.0 prompt structure is subject, action, camera, environment, sound, and cut timing. For a 4-6 second clip, keep it to one shot. For 8-15 seconds, use labeled shots so the model knows where each cut belongs.

The order matters. For single-shot clips, write subject and action first, then camera movement, environment or lighting, sound, and style only if it is still needed. For multi-shot clips, open with an overall format or continuity note, then label each shot and give every shot one action, one camera move, and a sound cue if it matters.

Write It Like A Director

If you have used image models for a while, this takes a minute to unlearn. Image prompting rewards moodboards: lighting words, texture words, beautiful nouns. Video prompting has to decide what happens on frame 1, frame 40, and frame 120, and how the scene gets from one to the next. Your prompt has to behave more like a director's note than a poster description.

Compare a weak prompt, "cinematic girl in tokyo at night, beautiful lighting, masterpiece," with this one:

Shot 1: A woman in a red coat steps into a rain-soaked Tokyo side street, neon reflections breaking across the pavement. The camera tracks beside her at shoulder height. Wet footsteps and distant traffic fill the background.

The second prompt is not better because it is longer. Every phrase has a job: the rain-soaked street gives texture, the neon reflections give light something to move across, the tracking camera gives motion, and the footsteps give timing.

Sound Cues That Help

Seedance generates synchronized audio, so sound cues are not decoration. They tell the model when events happen: the cap click lands when the bottle opens, the splash lands when the shoe hits water. The best cues are small and physical.

Avoid "great music," "epic sound," or "cinematic audio." They do not tell the model what event to sync with. If you need music timing, describe the tempo, beat, or transition point instead.

Seedance 2.0 Settings At A Glance

Draft In Fast, Finish In Standard

Seedance 2.0 Fast is for cheaper iteration; standard Seedance 2.0 is for final output, especially at 1080p. The prompt structure is the same, but Fast works better with fewer shots and simpler movement.

Prompts rarely land on the first try, so prove the movement first and raise quality second. Starting at final quality means spending more credits discovering basic prompt mistakes. For the full cost and quality breakdown, see Seedance 2.0 Fast vs Standard.

Seedance 2.0 Text-to-Video

Use Seedance 2.0 text-to-video when you have no image, product frame, or character reference and want to explore a scene from scratch. Write it as one shot a camera operator could film, match the number of beats to the duration, and name one physical camera move per shot. The tradeoff is identity stability: if the subject must stay exact, switch to image-to-video or references.

The biggest mistake is treating text-to-video like a search box. "A beautiful cinematic ad for a smartwatch" sounds clear to a human, but it does not say what happens on screen. A camera operator needs more: the watch sits on black glass, the second hand starts moving, cyan light catches the bezel, the camera slowly orbits at table height.

The Single-Shot Formula

[Subject] [does one action] in [specific setting], [camera movement], [lighting or atmosphere], [sound cue].

Example:

A young chef slides a copper pan across a stainless counter, low side-tracking camera, warm kitchen light, pan scrape and soft sizzle.

Once the single shot works, extend it into labeled shots for 8-15 second clips. Each shot should be a camera instruction with a starting frame, a motion beat, or a payoff. Written as one long paragraph, the same idea forces the model to decide where the cuts go.

Pace The Clip To Its Duration

Duration is a storytelling budget, not just a quality setting.

Ask for a full ad, a costume change, an orbit, a product reveal, and a reaction shot in 5 seconds and the model compresses everything until the clip feels confused. If the output feels chaotic, reduce either the number of shots or the number of actions per shot.

Camera Language That Works

Seedance responds better to physical camera directions than to generic style words.

Camera language is also how you control emotion. A locked-off wide feels observational, a handheld follow feels immediate, a slow push-in feels intimate, and a low-angle dolly feels powerful. Use one main camera move per shot; "orbit, dolly, zoom, handheld, drone" in one shot usually produces confused motion.

Compose For The Aspect Ratio

Start the prompt with the ratio, then compose for it.

When To Switch Workflows

Text-to-video is for mood tests, concept boards, generic social clips, landscape and b-roll ideas, and rough ad beats. Exact products, repeat characters, branded objects, visual continuity, and audio-synced edits belong in image or reference workflows. If you keep rerolling a text prompt because the product changes shape or the character looks different each time, that is not a prompting failure. It is a workflow signal.

A Better First-Draft Workflow

1. Write the simplest possible one-shot version.
2. Generate it short, usually 4-6 seconds.
3. Keep the useful motion and remove the rest.
4. Add shot labels only after the single-shot version works.
5. Move to Standard quality only after the prompt is stable.

The first draft is not the final clip. It shows you what Seedance understands from your wording. The prompt categories below are all text-to-video starting points built this way.

Cinematic Seedance Prompts

These work best at 8-15 seconds on the standard model when you want stronger resolution and more room for cuts. Cinematic prompts need an arc. Even a short clip should feel like it moves from setup to reveal, pressure to release, stillness to motion.

Rain Courier

What it does: a moody night-delivery shot with a simple three-beat structure. The rain, neon, and wheel splash give Seedance physical material to animate instead of just a vague "cyberpunk" look.

Shot 1: A lone courier rides a bicycle through a rain-soaked alley at night, neon signs reflecting in the puddles, handheld camera following from behind.
Shot 2: Low side-tracking shot as the wheel cuts through water, spray hitting the lens.
Shot 3: The courier stops under a red lantern and looks up, traffic hum and rain filling the street.

Best settings: 16:9, 8-10 seconds, 720p for drafts, 1080p for final.

Why it works: each shot has one visual task. The first establishes motion, the second gives tactile detail, and the third lands on a quiet pause. Seedance tends to do better with this rhythm than with three unrelated action beats.

Desert Runner Reveal

A desert runner crests a dune at sunrise, wind pulling sand across the frame, wide anamorphic composition, slow drone pullback, low ambient wind.

Best settings: 21:9 or 16:9, 6-8 seconds.

Why it works: the subject is small but readable, the camera move is slow, and the environment has one clear motion source: wind. This is a good prompt when you want atmosphere without making the model juggle characters.

Chef Plating Sequence

Shot 1: A chef places a single white plate under a warm pass light.
Shot 2: Close-up as sauce falls in a clean spiral, the camera pushes in slowly.
Shot 3: Steam rises from the dish as the kitchen noise fades into a soft room tone.

Best settings: 16:9, 8-10 seconds.

Why it works: food prompts succeed when they are about texture and timing. The spiral sauce, steam, and pass light are concrete enough to produce a premium restaurant feel without asking for too much.

Parking Garage Turn

A vintage car turns into an empty parking garage, headlights sweeping across concrete pillars, low front-mounted camera, tire echo and engine rumble.

Best settings: 16:9, 6-8 seconds.

Why it works: headlights are doing half the visual work. Seedance can use the moving light to make a simple turn feel cinematic.

Storm Skybridge

A woman in a structured red coat crosses a glass skybridge during a storm, lightning flashes outside, locked-off symmetrical frame, distant thunder.

Best settings: 16:9 or 9:16, 6-8 seconds.

Why it works: the locked frame prevents the prompt from becoming visually busy. The storm adds motion without forcing complex body choreography.

Product Video Prompts

Product prompts should preserve shape, material, and brand readability. Keep the motion simple unless the product is already familiar.

The trick with product clips is to resist the temptation to make the product do too much. A bottle does not need to fly through three rooms. It needs light, surface, texture, and a camera move that makes the object desirable.

Ceramic Mug

A matte ceramic coffee mug rotates slowly on a walnut table, morning light sliding across the curved surface, steam rising, soft spoon clink.

Best settings: 1:1 or 16:9, 6-8 seconds.

Why it works: the prompt gives the mug three gentle motions: rotation, light movement, and steam. That creates a premium loop without changing the product.

Headphone Case Reveal

Shot 1: A black wireless headphone case opens on a brushed metal table.
Shot 2: The earbuds lift slightly as blue LED reflections move across the surface.
Shot 3: Tight macro shot of the texture, clean studio silence with a soft electronic tone.

Best settings: 16:9, 8-10 seconds.

Why it works: unboxing and reveal formats are reliable because the action is easy to understand. The LED reflection gives the second shot a reason to exist.

Skincare Bottle

A skincare bottle stands on wet stone beside eucalyptus leaves, slow vertical dolly from label to cap, water droplets rolling down the glass.

Best settings: 9:16 for ads, 1:1 for feed posts.

Why it works: the vertical dolly is perfect for bottles because it turns label reading into the motion of the clip.

Running Shoe Splash

A running shoe hits a rain puddle in slow motion, low camera beside the sole, water splash frozen for a beat, city lights behind it.

Best settings: 16:9 or 9:16, 4-6 seconds.

Why it works: the splash hides small motion errors and makes the clip feel expensive. This is one of the easiest product formats to adapt.

Watch Macro

A stainless steel watch lies on black slate, the second hand starts moving, camera orbit at table height, subtle mechanical tick.

Best settings: 1:1, 6-8 seconds.

Why it works: the second hand gives the model a precise motion cue. The orbit and tick make the product feel alive without asking it to transform.

UGC And Ad Prompts

Use Seedance for short performance ads when you need a natural action beat, not a perfect talking-head script.

UGC prompts should feel handheld and slightly imperfect. Do not over-polish them. The point is to create a believable creator moment that can sit inside an ad account or TikTok feed without feeling like a glossy brand film.

Desk Lamp Before/After

Shot 1: A creator holds a small desk lamp in a messy home office, looking tired.
Shot 2: She switches it on and the room becomes warmer and cleaner.
Shot 3: Close-up of her smiling as the desk surface brightens, soft click and room tone.

Best settings: 9:16, 8-10 seconds.

Why it works: this uses the classic ad structure: problem, product action, payoff. The switch click gives Seedance a clean timing anchor.

Fitness Bag Moment

A fitness creator drops a gym bag on a bench, pulls out a shaker bottle, quick handheld camera, bright locker-room lighting, zipper and bottle snap sounds.

Best settings: 9:16, 4-6 seconds.

Why it works: it feels like a real phone-shot clip. The motion is simple and the sound cues are all tied to visible actions.

Cafe Study Reset

Shot 1: A student opens a laptop at a crowded cafe, overwhelmed by notes.
Shot 2: The screen glow reflects on their face as the workspace becomes organized.
Shot 3: They lean back and smile, cafe noise soft behind them.

Best settings: 9:16 or 1:1, 8-10 seconds.

Why it works: the transformation is emotional rather than literal. The viewer can understand the "before and after" without text on screen.

Home Cook Steam Shot

A home cook lifts a lid from a pan, steam rolls toward camera, handheld phone-style framing, natural kitchen light, sizzling sound.

Best settings: 9:16, 4-6 seconds.

Why it works: steam and sizzling make the clip sensory. This is the kind of prompt that feels better with native audio enabled.

Small Business Packing Clip

A small business owner seals a cardboard package with branded tape, close handheld shot, tape pull and box tap sound, warm warehouse light.

Best settings: 9:16, 4-6 seconds.

Why it works: it is specific enough to feel real but generic enough to adapt for almost any ecommerce brand.

TikTok, Reels, And Shorts Prompts

For vertical clips, tell the model the frame is vertical and keep the subject centered. Morphed's TikTok video guide covers more platform-specific workflows.

Short-form prompts need to respect the phone screen. A beautiful wide shot can collapse into a tiny, unreadable subject in 9:16. Keep the action close to camera, use one visual hook in the first second, and give the ending a loopable pose or impact.

Outfit Beat Change

VERTICAL 9:16. A creator steps from a dim hallway into bright sunlight, outfit changes on the beat, camera tracks backward, quick confident energy.

Best settings: 9:16, 6-8 seconds.

Why it works: the transition is tied to the beat and the lighting change. Seedance has a clear before/after moment to hit.

Notification Snap Zoom

VERTICAL 9:16. A phone on a desk lights up with a notification, the camera snap-zooms in, background room blurs, soft buzz sound.

Best settings: 9:16, 4 seconds.

Why it works: simple object motion plus camera motion. This is useful for app launches, newsletter hooks, and SaaS product teasers.

Drink Splash Macro

VERTICAL 9:16. A drink can drops into ice water, splash rises toward camera, fast macro shot, crisp fizz and ice crackle.

Best settings: 9:16, 4-6 seconds.

Why it works: the splash creates instant motion in frame. It is strong for thumbnails because the subject fills the vertical crop.

Fashion Turn

VERTICAL 9:16. A fashion model turns in front of a plain wall, fabric catches the light, smooth orbit camera, clean studio rhythm.

Best settings: 9:16, 6-8 seconds.

Why it works: a plain background protects the outfit. The fabric movement makes the clip feel like video instead of a moving photo.

Train Window Travel Shot

VERTICAL 9:16. A travel creator opens a train window as mountains pass outside, handheld follow shot, wind and rail sound.

Best settings: 9:16, 6-8 seconds.

Why it works: the moving background gives energy even if the person is mostly still. It is a strong travel-reel format.

Seedance 2.0 Image-to-Video

Seedance 2.0 image-to-video works best when the uploaded image owns the look and the prompt owns the movement. Tell Seedance what the image is responsible for, then describe one action, one camera move, any environment change, and a sound cue. Keep clips short when identity matters: 4-6 seconds for one clean movement, 8-10 for a product reveal or portrait beat.

This is where many prompts go wrong. People upload a perfect source image, then ask Seedance to redesign the whole thing. The job is animation, not transformation. Your still image is the set; the prompt is the director walking onto it and saying: move the camera here, let the light shift, let the fabric breathe, do not change the product.

Check The Source Image First

Seedance can animate a weak image, but it cannot fully repair one while also preserving identity.

If the source image is already doing a lot, write a quieter prompt. If it is simple, use the prompt to add atmosphere.

Image-To-Video Settings

A still image can support a strong moving shot, but not ten unrelated scene changes. Draft a 4-6 second Fast version at 720p, keep the prompt structure that worked, strip style words that did nothing, then rerun the winner in Standard at the final aspect ratio.

The Three Formats That Work Most Often

These are the formats closest to how real commercial b-roll is shot, so start with them before trying more ambitious sequences.

How To Write The Prompt

Write image-to-video prompts in this order: image role, action, camera, environment change, sound, constraints. A prompt like "Use the uploaded image as the exact product identity. The product stays the same shape and color. The camera slowly pushes in..." gives the model motion and constraints. "Make a cinematic beautiful commercial from this image, premium, dramatic, high quality" gives it only taste words.

If you only remember one rule: never let the prompt fight the image. When the image is already specific, write less about appearance and more about movement.

For products, protect geometry first. If the bottle, shoe, watch, or package changes shape, the clip is unusable. Product clips usually look more expensive when the camera moves slowly and the light does the drama; fast camera moves make products feel cheap unless the brand is intentionally high-energy.

For portraits, the face is the most fragile part. A slight turn feels alive. A wide smile, sudden laugh, head whip, and camera orbit in the same 5-second prompt will often change the person's identity. Treat portraits like close-up acting direction and move one thing at a time.

When To Add An End Frame

Use an end frame when both the start and finish compositions matter. Skip it for simple motion.

End frames are especially useful for ads, where the last frame is often the thumbnail, CTA moment, or product lockup. If the ending matters more than the motion, give Seedance the final composition instead of hoping it lands there. For a model chooser beyond Seedance, the image to video AI page compares options.

The five prompts below apply these rules.

Product Orbit From A Still

Use the uploaded image as the exact product identity. The camera slowly orbits clockwise around the product, soft studio light moving across the surface, no shape change.

Best settings: 1:1 or 16:9, 6-8 seconds.

Why it works: "no shape change" is not magic, but it tells the model that product fidelity matters more than visual invention.

Portrait With Subtle Life

Use the uploaded portrait as the exact character reference. The subject turns slightly toward the window, hair moves in a light breeze, subtle blink, soft room tone.

Best settings: 9:16 or 1:1, 4-6 seconds.

Why it works: subtle portrait animation survives better than exaggerated expression changes. The face stays readable.

Food Steam Push-In

Animate the uploaded food photo with steam rising naturally, a slow push-in toward the center of the plate, warm restaurant ambience.

Best settings: 1:1, 4-6 seconds.

Why it works: steam is forgiving. It gives motion without forcing the food itself to deform.

Car Headlight Reveal

Use the uploaded car image as the exact vehicle. The headlights turn on, camera lowers toward the front grille, rain beads on the hood.

Best settings: 16:9, 6-8 seconds.

Why it works: the clip has three physical events: lights, camera drop, rain texture. The vehicle design can stay stable.

Interior Sunlight Move

Use the uploaded interior design image. Sunlight shifts across the floor, curtains move gently, camera glides forward through the room.

Best settings: 16:9, 6-8 seconds.

Why it works: interior clips look better when the room itself stays calm and only light/fabric/camera move.

Reference-To-Video Prompts

Reference prompts work when each file has one job. Seedance 2.0 reference-to-video can use images, video clips, and audio clips, but overlapping instructions reduce control.

The cleanest reference prompts read like a small production note. One file is cast. One file is choreography. One file is soundtrack. If you try to make every reference control identity, lighting, camera, motion, and sound at once, the model has no hierarchy.

• Image reference = identity, wardrobe, styling, composition anchor
• Video reference = camera movement, pacing, motion behavior
• Audio reference = beat, rhythm, sound texture, sync cue

Product Beat Sync

@Image1 defines the product shape and material. @Video1 defines the slow orbit camera movement. @Audio1 defines the beat timing.

Shot 1: The product from @Image1 sits on black glass as the camera begins the orbit from @Video1.
Shot 2: Light sweeps across the surface on the first beat from @Audio1.
Shot 3: Final macro close-up lands exactly on the last beat.

Best settings: 16:9, 8-10 seconds, audio enabled.

Why it works: the prompt assigns a hierarchy. Image controls product, video controls movement, audio controls timing.

Character And Palette Mix

@Image1 is the character identity and outfit. @Image2 is the color palette. Keep the face and clothing consistent.

Shot 1: The character from @Image1 walks through a narrow city street using the warm colors from @Image2.
Shot 2: The camera tracks beside them at shoulder height.
Shot 3: They stop at a storefront and look toward the reflection.

Best settings: 9:16 or 16:9, 8-10 seconds.

Why it works: the second image is not competing for identity. It only controls palette, which makes the instruction easier to follow.

Dance Rhythm Transfer

@Video1 defines the handheld camera rhythm. @Audio1 defines the percussion timing.

A dancer moves through a plain rehearsal room, following the camera energy of @Video1. Each turn lands on the percussion accents from @Audio1.

Best settings: 9:16, 6-8 seconds.

Why it works: a plain room removes background complexity, so the motion and rhythm can carry the clip.

Fast Fixes For Bad Outputs

If the output looks like a moving still, add a stronger verb: steps, turns, pours, opens, lands, pulls, drifts. If the camera feels random, remove extra style words and name one camera move. If the subject changes identity, switch to image-to-video or reference-to-video and tell the prompt which reference owns identity.

The last fix is counterintuitive: most bad iteration responds to a weak output by adding more words, when the answer is often fewer words and a clearer movement.

Seedance is also not the right fit for every job. If you need highly precise talking-head dialogue, test lip sync carefully before using it in production, and expect more retries on dense multi-person choreography in one shot.

Related Seedance Guides

• Seedance 2.5 prompts: 30-second takes, audio, and reference briefs
• Seedance copyright: what is actually happening
• Seedance 2.0 API guide
• Best text-to-video AI generators
• Seedance 2.0 Fast vs Standard
• Seedance 2.0 vs Kling
• Seedance 2.0 vs Veo