Seedance 2.0 Prompt Guide: 20+ Copy-Paste Examples That Actually Work

If you've spent any time generating AI videos, you've probably run into the usual frustrations - characters that randomly change appearance, movements that feel stiff and mechanical, or scene cuts that just don't flow. Even when the output is technically correct, it rarely has the narrative punch or polished look you're going for.

More often than not, the problem isn't the model itself. It's that we're still thinking in terms of describing a scene rather than directing one.

That's exactly what Seedance 2.0 changes. Rather than simply generating visuals from a text description, it introduces a more structured approach to prompting - one that puts you in the director's chair. Through natural language, you can now define camera movements, control lighting, and shape how characters behave on screen.

In this Seedance 2.0 prompting guide, we'll break down how to write effective Seedance 2.0 prompts, with 20+ ready-to-use templates and practical Seedance 2.0 prompt tips to help you create more cinematic, emotionally resonant AI videos.

What Is Seedance 2.0?

Seedance 2.0 is a multimodal AI video generation model officially released in early 2026. Rather than just producing images, it transforms creative ideas into short-form videos with cinematic language and narrative rhythm.

Unlike traditional video generation models, Seedance 2.0 functions more like a complete AI-powered audiovisual production system. Whether you're starting from a simple visual concept or building a more complex story, it renders your ideas into professional-quality video in a natural, seamless way.

Key Capabilities of Seedance 2.0

1. Multimodal Support 🚀

Seedance 2.0 breaks creative barriers by simultaneously processing text, up to 9 reference images, 3 reference videos, and 3 audio tracks. Its exclusive @reference system lets you incorporate up to 12 external assets in a single prompt.

That means you can supply a character's face, action reference footage, and background music all at once - giving the AI everything it needs to fully understand and execute your creative vision.

2. Director-Level Precision Control 🎬

With Seedance 2.0, you can direct every detail using plain language: specify camera moves (pans, tracking shots, orbits), subtle facial expressions, body gestures, scene lighting and atmosphere, even edit pacing and transitions. The creation experience feels remarkably close to standing behind a real camera.

3. Native Audio-Video Sync 🎵

No post-production syncing needed. The AI generates a complete video with audio built in. Dialogue aligns perfectly with character lip movements, and subtle details like eye blinks or head nods sync automatically. Ambient effects and background music integrate naturally - one click, and you have a publish-ready video.

4. Advanced Physical Realism ✨

Built on the latest DiT architecture, Seedance 2.0 simulates the physical world with greater accuracy. Hand details are more stable, fabric movement looks more natural, and gravity and collision physics align more closely with real-world logic - all of which adds believability to the final output.

5. Advanced Natural Language Understanding 📝

Seedance 2.0 understands instructions written the way a real director would speak them. No need for complex technical jargon - you can describe scenes and pacing intuitively, which lowers the learning curve and keeps the creative process flowing.

Because of all this, a Seedance 2.0 prompt is no longer just a visual description - it's a true directing script. Let's look at how to write one.

The 3 Core Seedance 2.0 Prompt Structures

In Seedance 2.0, a well-structured prompt can mean the difference between a flat, generic clip and a cinematic-quality video. After extensive testing and community input, three core prompt frameworks have emerged that cover roughly 90% of use cases.

Here's the logic behind each one, along with practical examples and step-by-step instructions.

1. The Five-Segment Structure (Best for Beginners)

✅ Core logic: Subject + Scene/Atmosphere + Action/Performance + Camera Movement + Style/Lighting

This framework is easy to remember and works well for single-shot videos or simple compositions. It gives your Seedance 2 prompt a clear internal logic and reduces the chance of the model misinterpreting your intent.

🎯 Seedance 2.0 Prompt Example: Man with a red umbrella on a rainy night (text-only generation)

A serious man in a black overcoat, expression firm but tinged with melancholy. He slowly opens a red umbrella as raindrops slide along its edge. He stands on a neon-lit urban street; water splashes around him. The camera performs a slow push from a wide shot to a medium shot. Strong cinematic style, film grain, teal-orange color grading, 4K ultra HD, realistic physical simulation.

❓ How to use it:

Step 1. Copy the prompt directly into the Seedance 2.0 input field.

Step 2. Select default parameters: 10–15 second duration, 1080p resolution.

Step 3. Click Generate to preview the video.

Step 4. If the result isn't what you expected, fine-tune the style or lighting section to reinforce the atmosphere and regenerate.

2. The CRAFT Multimodal Framework (Best for Multi-Asset Projects)

✅ Core logic: Context + Reference (@assets) + Action + Framing/Timing + Tone/Audio

The CRAFT framework is especially useful when you're working with multiple references - images, video clips, or audio tracks. Use @Image to lock in a character's appearance, @Video to define movement style, and @Audio to reinforce atmosphere. This reduces the disconnect that often happens between reference assets and the generated animation, resulting in stronger aesthetic and narrative coherence.

🎯 Seedance 2.0 Prompt Example: Film Noir Detective Scene

Context: 1940s film noir detective office, heavy with smoke. Reference: Character face and wardrobe strictly follow @Image1; movement pacing is based on @Video1. Action: The detective steps out of the shadows, moves to the desk, and picks up a photograph, studying it with a grave expression. Framing: Dutch Angle composition with a tracking camera shot. Tone: Tense and melancholic atmosphere, classic film grain, background music @Audio1.

❓ How to use it:

Step 1. Upload your reference images or videos (total duration under 15 seconds).

Step 2. Incorporate them into the prompt using tags like @Image1 or @Video1.

Step 3. Organize your instructions following the CRAFT structure, and enable multimodal mode when generating.

Keep style intensity at a moderate level to balance the weight of reference assets with the model's creative output.

3. Timeline Storyboard (Best for Multi-Shot Narratives)

✅ Core logic: Break the video into specific time segments (e.g., 0–4s, 4–9s) and describe what happens in each.

This approach works especially well for longer videos with a narrative arc. It lets the model "edit" according to a predefined script, maintaining visual consistency and smooth pacing.

🎯 Seedance 2.0 Prompt Example: Epic Samurai Duel (multi-shot narrative)

0–4s: Wide shot. The samurai walks forward from a distance through a bamboo forest; wind ripples his clothing; morning mist fills the scene. Style: @Image1. 4–9s: Medium tracking shot. He draws his sword and takes his opening stance; leaves fall around him. 9–13s: Close-up. The blade cuts through the air; slow-motion water splashes. 13–15s: Whip pan. Sword flashes, Japanese-inspired epic atmosphere.

❓ How to use it:

Step 1. Set the total duration (10–15 seconds) and divide it into logical segments (e.g., 0–4s, 4–9s).

Step 2. Write each segment as an independent line describing action, character, and camera, keeping continuity between segments.

Step 3. If you have references, integrate them where they add value: @Image1 for appearance, @Video1 for pacing.

Step 4. Copy the full prompt, enable multi-shot optimization (if available), set ~15s duration, and generate.

Step 5. Review transitions; if they don't flow naturally, add "maintain narrative continuity" and regenerate.

Quick Reference - Which Framework to Use:

With the right framework, your Seedance 2.0 prompts stop being plain descriptions and become genuine directing scripts - giving your AI videos stronger narrative coherence, a more professional feel, and greater visual impact.

20+ Best Seedance 2.0 Prompt Examples (Copy & Paste)

To help you break through creative blocks and generate professional-level videos fast, we've compiled over 20 ready-to-use Seedance 2.0 prompt templates. They cover the most popular formats - from cinematic trailers and dance videos to product ads and digital avatar presentations - making full use of Seedance 2.0's physical simulation, lighting control, and timeline management capabilities.

Simply copy the Seedance 2.0 prompt you want, replace the asset references (@Image1, @Video1) with your own files, and let the AI bring your ideas to life.

Seedance 2.0 Prompts for Cinematic Trailers

1. Post-Apocalyptic Survivor 🏜️

Wind whips sand across the ruins of a collapsed city. A 28-year-old female survivor stands on a crumbling structure. Her torn gray cape billows in the wind; her face is covered in dust and old scars. At first she tilts her head against the gust. After a brief pause, something shifts on the horizon. In the distance, the blurred silhouette of a mechanical creature moves slowly. She raises her gaze. Her breathing steadies. Her eyes shift from confusion to resolve. The camera pushes from a medium shot to a close-up, guided by her line of sight - as if the world is coming back into focus. As the camera advances, the sandstorm softens and light breaks through the dust. Post-apocalyptic sci-fi style, desaturated earth tones. Realistic wind physics, light diffusion, and fabric movement.

2. Retired Soldier in the Rain 🌧️

Heavy rain hammers the roof of an abandoned factory. A hard-faced man in his early 30s stands at the edge. His worn dark-gray military coat clings to his soaked frame. A faint scar on his right cheek flickers under the distant glow of a neon sign. He looks down, lighting his last cigarette with a slightly trembling hand. A two-second pause. The smoke barely rises before the rain scatters it. In the distance, neon reflections shimmer in the puddles. He slowly raises his head - his expression shifting from exhaustion to cold resolve. The camera drifts slowly from a rear-side shot to a frontal medium shot, as if pulled by an invisible force. As it moves in, water streams down his face and his breath becomes visible. Dark realism style, high-contrast cool blue palette, cinematic film texture. Realistic rain physics, wet fabric clinging to the body, smoke dispersed by wind.

3. Cyberpunk Assassin 🤖

On a rain-soaked skyscraper rooftop, a 27-year-old cyberpunk assassin crouches on one knee. Her short silver hair is plastered to her face. Her left eye - a red cybernetic implant - emits a faint glow. She holds a luminous dagger that reflects the city's neon lights. The cybernetic eye activates a subtle scan. She slowly raises her head, locking onto a target. The camera begins a 360° orbit driven by the city's shifting light, as if the environment itself is analyzing her. During the rotation, neon reflections glide across wet metal and her suit's surface. The motion stops at a low frontal angle. The eye's glow intensifies. Cyberpunk aesthetic, high-contrast purple and red neon. Realistic rain reflections on metal, natural movement of wet fabric.

4. Gunslinger at Dusk 🤠

In the desert at sunset, a light wind stirs fine sand. A 42-year-old cowboy stands motionless. The brim of his hat shadows his eyes. His gray beard sways in the wind. At first - no movement. The wind picks up in the distance. The moment the sun fully disappears below the horizon, he slowly raises his hand and draws his weapon. The camera pushes from a medium shot to a chest-level close-up, guided by the light shift - the world enters the duel moment. As it advances, dust grows more visible and shadows lengthen. Classic Western aesthetic, warm orange tones, film grain. Realistic sand dynamics, natural light transition.

5. Steampunk Inventor ⚙️

On the deck of a massive airship, gears turn slowly as steam vents in rhythmic bursts. A 29-year-old inventor stands at the center. Copper goggles rest on her forehead. Her curly brown hair is tossed by the wind. She adjusts a mechanical device, eyes down. A pause. Suddenly the steam intensifies. She looks up, as if sensing a force about to engage. The camera moves from a side angle to a frontal shot, driven by the mechanical rhythm - gears accelerating. Behind her, the cloud sea is pierced by columns of steam, and mechanical shadows sweep through the space. Steampunk aesthetic, warm copper and gold tones. Realistic steam expansion, mechanical inertia, worn fabric textures.

Seedance 2.0 Prompts for Digital Avatar Explainer Videos

1. Professional Speaker 🎤

A male speaker in his early 30s, wearing a navy suit with a clean short hairstyle, projects confidence and approachability. Frontal medium shot inside a modern glass-wall office, with a softly blurred urban skyline in the background. A gentle push-in camera move. He pauses briefly, takes a soft breath, and begins to speak: "AI is redefining the way we live." Lip-sync aligns with @Audio1. His hands accompany the speech with natural, measured gestures; he occasionally nods or offers a slight smile. Consistent eye contact with the camera throughout. Soft cinematic lighting defines the face with warm neutral tones. Breathing is visible, and gestures follow speech flow without feeling mechanical or repetitive.

2. Animated Cartoon Girl 🎀

A cartoon-style girl with pink pigtails and large round eyes sits in a rainbow-colored animated room. Medium shot with a slight camera sway for energy. She blinks and leans slightly forward, then says with enthusiasm: "This little trick is super useful!" Lip-sync aligns with @Audio1. Expressions are lively but fluid, with gestures shifting to match the vocal rhythm. The head tilts occasionally for warmth. Lighting is warm, and the environment is bright and saturated. Expressions and movements evolve with the speech, avoiding repetitive loops.

3. On-Camera Host 📺

A polished on-camera host with long straight black hair, red lip makeup, and a white blouse. Medium-close frontal shot inside a high-end studio set. Three-point lighting creates volume and facial definition. The camera advances slowly and steadily. She adjusts her posture slightly and begins speaking with poise: "This product will transform your daily efficiency." Precise lip-sync with @Audio1. Gestures are subtle and refined, naturally complementing the speech. She maintains firm eye contact with the camera, projecting professionalism. The background is softly blurred with a clean, premium aesthetic.

4. Veteran Historian 📚

An elderly historian with white hair, round glasses, and a gray sweater. Medium shot inside an ancient library lit by flickering candlelight. The camera slowly moves in. He first looks down in quiet reflection, then raises his eyes and says: "A thousand years ago, that battle…" Precise lip-sync with @Audio1. Gestures are measured and restrained; he nods occasionally to emphasize a point. His expression conveys wisdom and calm, with an unhurried speech rhythm. The atmosphere is enhanced with warm tones and a subtle cinematic grain that reinforces the historical character.

Seedance 2.0 Prompts for Dance Videos

1. Latin Dance 💃

A Latin dance couple in a nocturnal tropical setting - she in a red dress, he in a black shirt. Warm golden lighting creates a samba atmosphere, with natural dress movement. 0–5s: Close spin at slow tempo; the dress flows softly. 5–10s: The rhythm of @Audio1 intensifies - fast spins enter, with a leg lift and an exchange of glances. 10–13s: Crossed-step sequence at a denser beat. 13–15s: Musical pause and final pose in an embrace. The camera transitions from a circular medium shot to a gradual close-up, synchronizing footwork and framing throughout.

2. Neon Street Dance 🕺

A street dancer based on the look of @Image1, wearing a black hoodie, on a rainy night street lit by neon. The background shows a blurred crowd, and the wet ground reflects the lights. 0–3s: Subtle warm-up movement, shoulders following the beat. 3–7s: The beat from @Audio1 drops - footwork and jumps. 7–10s: Rhythm intensifies, fast spin and landing. 10–15s: On the beat drop, a final freeze. The camera mirrors the music: handheld tracking at the start → whip pan on accents → slow push for the closing. Color particles burst on the beat hits. Maintain character consistency, perfect music sync, realistic physics, and cinematic lighting with a strong sense of rhythm.

3. Cyber Electronic Dance 🤖

A dancer floats in a cyberspace environment, movements fully synced to the electronic beats of @Audio1. 0–4s: Light floating; the body vibrates with low frequencies. 4–8s: Beat starts - segmented popping moves. 8–12s: Acrobatic jump on the heavy hit. 12–15s: Suspended in the air during the beat drop. Quick camera cuts with whip pans on accents. Neon lights pulse with the rhythm, particle explosions accompany the music, and the virtual grid in the background subtly distorts.

Seedance 2.0 Prompts for Action Videos

1. Cyber City Chase 🏃‍♂️

A character following the reference of @Video1 runs at high speed through rainy neon city streets, with minor visible injuries. 0–4s: Wide tracking shot - fast but grounded steps, realistic splashes through puddles. 4–9s: Slow motion as the character leaps over obstacles; prosthetic leg grazes the ground, generating mechanical sparks. 9–13s: Sharp turn, neon lights reflecting dynamically on wet pavement. 13–15s: Character stops, looks back - breathing heavy but alert, rain streaming down the face. Environment: realistic rain droplet gravity, shoe-pressure feedback, neon reflections with delay. Camera rhythm: tracking → slow-motion jump → circling turn → tense close.

2. Swordsman Duel ⚔️

Two swordsmen based on @Video1, standing in a forest clearing, facing each other. Wind lifts slowly spinning leaves, building a tense atmosphere. 0–5s: Static medium shot - held breaths, eyes scanning for weakness. Sleeves and leaves move with the wind, creating dynamic tension. 5–10s: The clash erupts suddenly. Fast camera with push-pull following the rhythm of strikes; metal clangs spark realistically; slow-motion blood droplets fly and fall under gravity. 10–15s: Camera circles the victor. The opponent falls; the winner pauses to confirm the outcome and sheathes the sword. Dust settles slowly. Physics: metal impact, blood trajectory, clothing inertia, airborne leaf dynamics. Camera design: static → rapid push-pull → stable circular close.

3. Futuristic Hand-to-Hand Combat 🤖

Two futuristic warriors based on @Video1 engage in close-quarters combat in a high-tech arena. 0–5s: Medium standoff shot - assessing weaknesses; mechanical arms charge up energy gradually. 5–10s: High-speed exchange of blows; contact points generate electric arcs with real physical feedback. 10–15s: Slow motion of a blow that sends the opponent down. Armor fragments fly; the fallen warrior tries to recover. Visual style: high-contrast blue-purple lighting, sci-fi energy particles, armor with realistic physics. Camera design: static tension → rapid conflict → slow-motion explosion.

4. Explosive Escape 💥

A tactical soldier based on @Image1, covered in dust, wearing a tactical vest with minor wounds - cold expression but visibly tense. 0–4s: Advances cautiously through a burning industrial zone, constantly scanning for threats. 4–8s: A lateral explosion - instinctively ducks, protecting his head from the blast. 8–12s: Slow motion moving through fire and flying debris, gaze focused. 12–15s: Exits the burning area, stops and looks back - only when he confirms safety does he allow himself to breathe. Motivation: pure survival instinct, raw realism (no heroic poses). Camera logic: advance → explosion → crossing → safety confirmation.

Seedance 2.0 Prompts for Promotional Videos

1. Fashion Runway 👗

A model wearing a long black silk dress based on @Image1, on a high-end runway stage. 0–4s: Side-angle tracking shot following her steps; the dress moves naturally with her body's center of gravity. 4–8s: The camera gradually orbits the model, showcasing the fabric's sheen under the lights as it flows smoothly. 8–12s: The model holds a confident gaze and steady stride; when she turns, the skirt forms an elegant curve. 12–15s: The camera pushes in to a frontal close-up, landing on a fixed frame. Environment: softened runway lights, blurred audience silhouettes. Visual style: elegant, understated, confident. Physical details: natural silk inertia, light following movement.

2. Smartphone Product Video 📱

A premium smartphone based on the look of @Image1, with a metallic body and glass edges that softly catch light in a diffused studio environment. 0–3s: The product floats against a solid-color gradient background, slowly rotating 360° to reveal the edges and material details. 3–7s: Macro shot drifting to the phone's side panel - light glides across the metallic surface, highlighting manufacturing precision. 7–10s: The screen gently illuminates, revealing an animated fingerprint sensor. 10–15s: The camera slowly drifts into the center of the screen, where UI elements breathe subtly. Environment: minimalist tech aesthetic, premium and futuristic feel. Visual focus: realistic metallic reflections, glass refraction, smooth light transitions on screen.

3. Coffee Machine Promo ☕

A coffee machine placed on a wooden table based on the look of @Image1, in a warm and quiet atmosphere. 0–4s: The machine powers on; steam rises slowly. 4–8s: Macro shot of coffee drops falling into the cup, creating natural ripples. 8–12s: Steam envelops the cup, visually suggesting aroma. 12–15s: The camera drifts gently toward the coffee's surface, reflecting a fine crema layer. Environment: morning home atmosphere, warm wood reflections. Realism: steam dispersion, liquid gravity, natural ripples in the cup.

Seedance 2.0 Prompts for Everyday Life Videos

1. Rainy Day Home Vlog 🏠

A 20-year-old in an oversized sweater, curled up in the corner of a couch reading a book, while rain slides slowly down the windowpane. 0–5s: Wide shot of the living room, rain sound in the background. 5–10s: The camera gently pushes into a medium shot; she turns a page and smiles softly. 10–15s: Close-up of raindrops sliding down the glass, while steam rises from a mug of hot cocoa simultaneously. Environment: relaxed, safe, slow rhythm. Realism: water flow, steam dispersion, slight page movement.

2. Japanese Sushi 🍣

A delicate Japanese sushi spread arranged on a wooden tray - salmon glistening softly, accompanied by a bowl of miso soup with steam rising slowly. 0–4s: Wide overhead shot; a hand enters the frame gently to adjust the chopsticks. 4–8s: Chopsticks pick up a piece of sushi, pausing briefly mid-air with a natural wrist adjustment. 8–12s: Lightly dipping it in soy sauce, creating subtle ripples on the liquid's surface. 12–15s: Chopsticks exit the frame; the soup shifts gently and steam continues to rise. Highlighted actions: small hand pauses, the dipping gesture, soft utensil contact. Realism: soy sauce surface tension, steam dispersion, natural ingredient inertia.

Practical Seedance 2.0 Prompt Tips

While the templates above make it easy to get started, truly precise control over AI video generation comes from understanding how to fine-tune each instruction. Small changes to a Seedance 2.0 prompt can be the difference between a visually flat clip and one with genuine cinematic quality. Here are four advanced Seedance 2.0 prompt tips to help you level up.

Tip 1: Understand the Multimodal Reference Priority System

The @reference system in Seedance 2.0 isn't just an asset pile - it follows a clear hierarchy:

• @Audio (rhythm anchor): Used for lip-sync and beat matching in edit timing.
• @Video (movement anchor): Transfers motion trajectories and camera language. The most effective way to replicate complex movements, like a Hitchcock zoom.
• @Image (visual anchor): Locks in character appearance - face (Face ID), wardrobe, and overall style. Use a mid-body portrait with a simple composition for best results.

Tip 2: Use Precise Cinematic Language

Incorporating film production terminology noticeably elevates the cinematic quality of the output:

• Special shots: POV (first person), 360 orbit (wraparound movement), Rack focus (focus shift)
• Pan / Tilt: Pan left (space exploration), Whip pan (dynamic transition)
• Track / Dolly: Tracking shot (subject follow), Truck left/right (lateral move)
• Push / Pull: Slow push in (emotional intensity), Dramatic pull out (environment reveal)

Tip 3: Keep Characters Consistent

Visual consistency remains one of the biggest challenges in AI video generation. To minimize variation, use a combined strategy:

• Explicit instruction: Add a line at the end of your prompt like: Maintain facial features and wardrobe fully consistent with @Image1.
• Text reinforcement: Even when you have a reference image, reiterate key traits in text (e.g., "short silver hair," "mole under left eye").
• Visual anchoring: Always use the same facial reference (@Image1) across all prompts in a project.

Tip 4: Balance Length, Detail, and Positive Framing

To ensure the model interprets your instructions correctly:

Optimal length: Keep your prompt between 120 and 280 words. Prompts that are too short tend to produce unpredictable results, while excessively long ones can dilute the model's focus.

Information hierarchy: The model generally prioritizes elements that appear early in the prompt. Structure your information like this:

Subject → Action → Camera → Style

In plain terms: who does what first, then how it's filmed, then visual treatment.

Positive framing: Seedance 2.0 responds better to affirmative instructions than negative ones. Rather than describing what to avoid, guide the output with terms like:

• Detailed hands, Anatomically correct → improved anatomy
• Physically accurate, Natural motion → better movement
• Ultra sharp, Masterpiece → higher visual quality

Conclusion

Seedance 2.0 represents a genuine turning point in AI video generation - not just as a technology leap, but as a creative breakthrough. Once you master Seedance 2.0 prompts, you stop being someone who "generates videos" and become a real director capable of building your own visual world.

⚠️ Important Note:

Seedance 2.0 is currently rolling out gradually and is only available in select countries and regions. If access isn't available in your area yet, don't let that slow your creative momentum:

• Seedance 1.5 Pro: A robust, capable version that handles the vast majority of video generation tasks with ease.
• SeaArt AI: A powerful and accessible platform that already integrates the most advanced video models on the market, including Kling 3.0 and Wan 2.6 - an ideal option for creating high-quality videos while you wait for Seedance 2.0.

The future of video is already here. The director's chair is waiting - and your creativity is ready for its close-up. Time to hit record.