Seedance 2.0 Prompts: Copy-Paste AI Video Examples

Structured prompts for ByteDance-style multimodal video workflows—click a block to copy, then adapt duration and @ references for your app.

March 29, 2026 10 min read AI Image Prompt Gallery

These templates follow the subject + action + camera + style + constraints pattern discussed in public Seedance 2.0 guides (example overview). For the full methodology, see our Seedance 2.0 prompt guide. Replace placeholder tags with your product’s syntax if it differs.

How to Use

Click any dark prompt box (or use Copy Prompt) to copy text. Keep one strong idea per clip; split complex stories into multiple generations and edit together.

Five-Part Cinematic Prompts

Single-clip scenes

Subject: Vintage bookbinder in a sunlit workshop, linen apron, silver hair tied back. Action: Threads a needle, sews a spine with steady hands, blows dust off the page edge. Camera: 50mm static medium close-up, slight rack focus from hands to face. Style: Warm tungsten key, dust motes, shallow depth of field, tactile paper texture. Constraints: 6 seconds, no text, no logos, realistic thread tension and paper bend.
Subject: Electric hatchback on a cliff road after storm. Action: Water sheets off the roof; wipers complete one slow cycle; driver visible as silhouette. Camera: Low angle tracking shot, 24mm, road spray toward lens. Style: Teal-orange grade, wet asphalt speculars, anamorphic flare streaks. Constraints: 8 seconds, consistent car silhouette, no license plate legibility.
Subject: Ballet dancer in a bare studio, morning side light. Action: Single pirouette, arms open, skirt hem follows two beats late. Camera: Slow dolly arc 90 degrees around subject, eye line locked to lens once. Style: High-key natural light, polished wood floor reflections, minimal aesthetic. Constraints: 7 seconds, one dancer only, no mirror duplication glitches.
Subject: Night market food stall, steam rising from a wok. Action: Chef tosses noodles; flame flares; crowd blur behind. Camera: Over-the-shoulder into wok, handheld micro-movement, 35mm feel. Style: Saturated practical neons, humid air, documentary energy. Constraints: 5 seconds, no readable signage, food physics believable.

Storyboard Scripts (Timecoded)

Multi-beat blocks

【Style】Premium tech product launch, clean white studio, soft reflections. 【Duration】10 seconds [00:00-00:03] Shot 1: Hero (360 spin platform). Matte black earbuds case rotates on a pedestal; soft gradient background; subtle contact shadow. [00:03-00:07] Shot 2: Macro (detail). Lid opens with precision hinge; glossy interior catchlight; faint brand-free silhouette only. [00:07-00:10] Shot 3: Lifestyle insert (hands). Hands lift one earbud; shallow DOF; calm confidence, no face required. Constraints: consistent product proportions; no readable text; realistic plastic and magnet snap.
【Style】Wildlife documentary, long-lens nature, neutral grade. 【Duration】12 seconds [00:00-00:04] Shot 1: Habitat (wide). Mist over a boreal forest river; distant bird calls; camera locked on tripod. [00:04-00:08] Shot 2: Subject (telephoto). Fox steps from tree line, breath fog, ears rotate toward off-screen sound. [00:08-00:12] Shot 3: Exit (pan follow). Fox trots out left; shallow snow kick; camera pans smoothly, no motion blur smear. Constraints: single fox; natural fur clumping; no human structures; no on-screen text.

Multimodal Starter Lines

Pair with your uploads

@Image1 as first frame: protagonist facing camera in a leather jacket. Action: slow turn to profile, hair reacts to wind from camera left. Camera: 85mm, subtle push-in. Style: neo-noir, sodium rim, fine grain. Constraints: 6 seconds, keep jacket color matched to @Image1, no subtitles.
@Image1 product hero on marble slab. @Video1 approximate camera path: slow top-down orbit tightening. Style: luxury tabletop, softbox key, crisp speculars. Constraints: 8 seconds, product scale locked to @Image1, no extra objects entering frame.
Interior car dialogue beat. @Audio1 as reference tone for male voice: calm, low register. Visual: driver at wheel, golden hour through windshield, hands at ten and two. Camera: passenger-side medium, stable mount. Constraints: 5 seconds, lip sync approximates @Audio1 rhythm, no dashboard text readable.

Product and Commercial

Packshots and lifestyle

Subject: Skincare bottle with pump, frosted glass, single botanical stem beside it. Action: Pump depresses once; pearl serum bead forms; stem shadow shifts slightly. Camera: Macro 100mm, locked tripod, symmetrical composition. Style: spa minimal, soft gradient backdrop, clean specular edge. Constraints: 5 seconds, label remains illegible, liquid viscosity realistic.
Subject: Running shoe on a treadmill belt, urban gym ambience. Action: Lacing tightens; foot lifts; belt motion blur at sole contact. Camera: Side profile low angle, 35mm, short gimbal move matching stride. Style: High-energy commercial, cool cyan fill, warm accent rim. Constraints: 6 seconds, single shoe model, no brand marks, consistent sole deformation.

Travel and Atmosphere

Wide and intimate

Subject: Desert highway straightaway at sunset, lone motorcyclist as small figure. Action: Bike moves toward horizon heat shimmer; dust trail thin and low. Camera: Extreme wide, slow crane-down from sky to road level. Style: Lawrence-inspired scale, long shadows, warm sand palette. Constraints: 10 seconds, no road signs readable, heat haze subtle not smeared.
Subject: Rainy Tokyo alley, vending machine practical as key light. Action: Commuter folds umbrella; condensation drips; steam from a vent grille. Camera: 40mm handheld, slow walk forward, foreground neon bokeh. Style: Cyber-noir humidity, wet pavement reflections, film grain light. Constraints: 8 seconds, no legible ads, crowd depth as soft bokeh only.

Dialogue Micro-Beats

Short spoken moments

Subject: Journalist in a radio booth, large diaphragm mic, foam walls. Action: Leans in, speaks one sentence, pulls back slightly on last word. Camera: Through glass control-room angle, slight reflection layering. Style: Intimate documentary, warm practicals, subtle head shadow on mic. Dialogue cue: "We are not recording silence—we are recording room tone." Constraints: 6 seconds, no on-mic logo readable, mouth timing matches line length.
Subject: Two friends on a rooftop at blue hour, city bokeh behind. Action: One hands a drink; clink glasses; genuine laugh without exaggerated mime. Camera: Two-shot medium, slow handheld drift, 40mm. Style: Indie romance palette, soft edge falloff, natural skin texture. Dialogue cue: A laughs: "You actually showed up." B: "Traffic was a movie." Constraints: 9 seconds, no subtitle burn-in, eyelines matched.

Conclusion

Use these blocks as scaffolding: swap subjects, tighten camera verbs, and add your own constraint list after you see where the model drifts. When a provider exposes multimodal inputs, attach references before you add more adjectives in text.

Next Step

Read the full Seedance 2.0 prompt guide for workflows, failure modes, and official model context links.

Featured AI Video Prompts