Seedance 2 Prompts und Beispiele

Seedance 2.0 gehört aktuell zu den besten KI-Videotechnologien für die Generierung von cinematischen und actionreiche Videos. Hier zeige ich dir einige Beispiel-Videos die ich damit generiert habe – inklusive Prompts.

Seedance 2.0 ohne Abo (Lifetime-Deal):

Seedance 2.0 Videos + Prompts

Folgende Videos habe ich mit Seedance 2.0 generiert. Die Prompts kannst du direkt kopieren.

Cinematisches Multi-Szenene-Video

Cinematic multi-scene short film, ultra-realistic, Arri Alexa Mini, 35mm lens. Full POV perspective throughout all scenes, seamless transitions between locations.

SCENE 1 — Tokyo, 5:47 AM
You step off a bullet train onto an empty platform. Steam rises from a vending machine beside you. Your reflection slides across the train window as it pulls away. You walk through the turnstile into a city that is just waking up, neon signs still glowing against a pale blue sky.

SCENE 2 — Moroccan Desert, Midday
Your boots crunch across burning orange sand. A camel caravan passes slowly to your left, bells ringing softly. You pull a worn map from your jacket pocket, look up at a crumbling sandstone fortress on the horizon, and start walking toward it.

SCENE 3 — Norwegian Fjord, Dusk
You sit at the bow of a small wooden boat cutting through black mirror water. Cliffs rise hundreds of meters on both sides, waterfalls catching the last light. Your hand trails in the freezing water. A single light glows in a cabin far above on the cliff face.

SCENE 4 — New York City, 2 AM
You push through the door of a tiny underground jazz club. The trumpet player locks eyes with you as you walk in. The room is packed, smoky, loud. You find a seat at the bar, and someone slides a glass toward you without being asked.

SCENE 5 — Somewhere Unknown, Dawn
You wake up in a tent. You unzip the door and crawl out into a meadow covered in low morning fog. Mountains in every direction. A fire is already burning beside the tent. A cup of coffee sits on a stone, still warm. You are completely alone but it feels like someone was just here.

Full POV throughout, ultra-realistic, cinematic color grading per scene, natural lighting, ambient sound design, seamless location transitions, emotional travel atmosphere, shallow DOF, no distortion, no stretching.

Video mit „Freeze-Effekt“

 Cinematic time-freeze short film, 15s, ultra-realistic, Arri Alexa Mini, 35mm lens. Lighting: Rain-soaked nighttime street carnival, flickering neon signs mixed with warm tungsten bulbs from food stalls, reflections on wet asphalt, drifting smoke from grills, deep shadows, shallow DOF, heavy atmospheric haze.

[0:00–0:03] A chaotic nighttime street carnival in the middle of the city. People are laughing, shouting, dancing, bumping into each other. A neon Ferris wheel spins in the background. A street performer accidentally knocks over a cart of glowing glass marbles, and a cyclist swerves through the crowd. Steadicam frontal medium shot tracking a calm man in jeans and a black shirt walking straight through the chaos. A paper lantern has just caught a gust of wind above him. He raises his hand and snaps.

[0:03–0:06] The snap triggers a spherical shockwave with visible light refraction, bending the neon reflections across the wet street. The entire carnival freezes instantly. Hundreds of glowing glass marbles hover in mid-air, each reflecting tiny upside-down neon signs. Raindrops hang suspended like crystal beads. A cloud of cotton candy is frozen mid-tear, stretching like pink smoke. Sparks from a food stall grill float motionless in the air. A cyclist is frozen sideways, wheels inches above the ground. Absolute silence.

[0:06–0:09] Only he moves, his sneakers splashing softly through shallow puddles. The camera retreats as he walks between frozen carnival-goers, ducks under a suspended paper lantern, and steps around a frozen burst of spilled soda hanging in the air. He reaches into the floating storm of glowing marbles and plucks one between two fingers.

[0:09–0:11] He stops in front of a frozen little arcade machine where a claw is about to miss a prize by a few centimeters. A child’s face beside it is frozen in total suspense. He tilts his head, smirks, and places the glowing marble perfectly beneath the claw, adjusting fate by the smallest possible detail.

[0:11–0:15] He snaps again, reversing the shockwave. The neon reflections ripple back into motion, marbles clatter across the wet asphalt, rain falls, cotton candy tears apart, sparks fly upward, the cyclist lands and swerves safely, and the carnival roar returns instantly. The claw drops, grabs the prize perfectly, and the child cheers as the man walks off-frame into the neon rain.

Action-Szene (wie in einem Hollywood-Blockbuster)

Cinematic Hollywood-scale disaster action short film, 15s, ultra-realistic, Arri Alexa 65, anamorphic 40mm lens. Lighting: Nighttime downtown boulevard during a sudden superstorm, cold blue lightning flashes mixed with warm orange streetlights and red emergency lights, wet asphalt reflections, heavy rain, wind-driven mist, deep shadows, shallow DOF, massive cinematic contrast.

[0:00–0:03] A crowded downtown boulevard at night, traffic frozen in panic but still moving slowly, rain hammering down, umbrellas flipping inside out, car horns blaring. A calm man in a soaked black coat steps into the street as everyone runs past him. In the distance, a giant construction crane begins to collapse between skyscrapers. Camera: wide anamorphic tracking shot pushing forward through rain and headlights.

[0:03–0:06] Lightning strikes the crane, showering sparks across the skyline. The crane arm tears loose and swings toward the boulevard, smashing through glass windows high above. The man looks up, expression controlled, as debris rains down around him. Camera whip-pans from the falling crane to his face, speed-ramping into slow motion.

[0:06–0:09] He sprints forward through traffic, sliding across the hood of a taxi as shattered glass and metal fragments explode behind him. A bus skids sideways, headlights cutting through rain. Camera switches to low-angle handheld tracking beside his feet, water splashing violently with every step, intense motion blur.

[0:09–0:12] The crane arm crashes into the street behind him, sending a shockwave through parked cars. He dives through the open door of a moving tram just before a wall of sparks rips across the road. Inside the tram, passengers gasp as he rolls to his feet, soaked and breathing hard. Camera follows in one continuous chaotic movement.

[0:12–0:15] The tram bursts out from beneath falling debris, racing through sheets of rain as the boulevard collapses behind it. The man steadies himself, looks back through the rear window at the burning skyline, then gives a small relieved smile. Final shot: lightning silhouettes the city as emergency sirens rise — hard cut to black.

Ultra-realistic Hollywood disaster action, massive scale, photorealistic rain, fire, sparks, collapsing crane, dynamic camera movement, anamorphic lens flares, speed ramps, intense motion blur, practical-feeling VFX, no distortion, no stretching.

Der Magier

Cinematic fantasy-heist short film, 10s, ultra-realistic, Arri Alexa Mini, 35mm lens. Lighting: Midnight luxury hotel lobby, golden chandeliers mixed with cold moonlight from tall glass doors, polished black marble reflections, drifting dust particles, deep shadows, shallow DOF, elegant high-contrast look.

[0:00–0:02] A sharply dressed elderly magician in a velvet burgundy suit walks calmly through a silent luxury hotel lobby after midnight. The camera glides beside him in a smooth tracking shot. Behind him, cleaning robots move quietly across the marble floor, and a grand piano plays by itself in the corner.

[0:02–0:04] He stops before a locked glass display case containing a glowing blue diamond. He removes a worn playing card from his sleeve and flicks it toward the case. The card spins through the air in slow motion, its edges shimmering with tiny sparks.

[0:04–0:06] The card touches the glass without breaking it. Instead, the entire display case ripples like water. The blue diamond liquefies into glowing light, flows through the glass, and reforms as a small crystal bird hovering above his palm.

[0:06–0:08] Suddenly, red laser security beams appear across the lobby. The magician smiles, raises one finger, and the crystal bird flies between the lasers, folding beams into glowing red ribbons. The camera follows the bird in a fast, elegant FPV shot between chandeliers and marble columns.

[0:08–0:10] The bird lands on his shoulder and turns back into the diamond. He places it inside his pocket, taps his cane once, and every chandelier flickers out except one spotlight above him. He bows politely to an empty lobby — hard cut to black.

Ultra-realistic, elegant magical realism, cinematic heist atmosphere, smooth camera movement, photorealistic reflections, subtle fantasy effects, no distortion, no stretching.

Audi RS6 – „Vienna Drift“

Cinematic nighttime street racing sequence in Vienna, a focused driver in a black Audi RS6 gripping the steering wheel, intense stare, Vienna city lights reflecting across the windshield, tension building before sudden acceleration.

Camera: Fast multi-angle system with seamless transitions, interior close-up → over-the-shoulder → exterior tracking → low ground-level shots, ultra-dynamic camera movement, whip-pans + speed-ramp transitions + motion-blur masking cuts, creating the illusion of one continuous flow.

(0–2s) Interior close-up of the driver inside the black Audi RS6, hand tightening around the gear shifter, subtle breathing, dashboard lights glowing, premium Audi interior details visible.

(2–4s) Over-the-shoulder shot, the road ahead stretching into Vienna’s neon-lit nighttime streets, wet asphalt reflecting streetlights and tram wires above, engine vibration building.

(4–6s) Extreme close-up of a finger pressing the boost / performance mode button, immediate ignition-like reaction, dashboard lights intensify.

(6–8s) Explosive acceleration, camera jumps to a side exterior tracking shot, the black Audi RS6 launches forward with a massive burst of speed, glossy bodywork reflecting Vienna’s historic façades and modern city lights.

(8–10s) Ultra-low ground-level shot close to the asphalt, wheels spinning at extreme speed, wet pavement, tram tracks, lane markings, and streetlights streaking past.

(10–12s) High-speed chase through narrow Vienna streets, sharp turns past elegant old buildings, tram lines, illuminated shopfronts, tunnel passages, and sweeping urban curves, camera whip-pans between angles, reflections and light trails intensify the sense of speed.

Dense Vienna urban night environment, wet asphalt reflecting neon signs, warm streetlights, tram tracks, historic architecture, tunnel passages, passing streetlights, high-speed city atmosphere.

Ultra-realistic, inspired by the energy of Fast and Furious, photorealistic lighting, intense motion blur, high-contrast neon reflections, cinematic depth of field, extreme sense of speed, smooth transitions, no distortion, no stretching.

Im Weltraum

Cinematic short film, ultra-realistic, Arri Alexa Mini, 35mm lens.

A lone astronaut floats outside a space station during a solar storm. The sun is erupting in slow waves of golden plasma, painting the entire scene in burning amber and deep black shadow. Earth glows silently below.

Something goes wrong. A tether snaps. The astronaut begins drifting away from the station, slowly at first, then faster. No panic. Just silence, heavy breathing inside the helmet, and the reflection of Earth in the visor.

With one last reach, fingertips brush the station's hull — and catch.

Ultra-realistic, cinematic space atmosphere, photorealistic Earth and sun, deep silence, emotional weight, anamorphic lens flares, no distortion, no stretching.

Muscle Car Action

Cinematic Hollywood sci-fi action short film, 15s, ultra-realistic, Arri Alexa 65, anamorphic 40mm lens. Lighting: Desert highway at sunset, burning orange horizon mixed with cold cyan energy glow, dust clouds, reflective car surfaces, harsh rim light, deep shadows, shallow DOF, epic cinematic contrast.

[0:00–0:03] A lone woman in a dusty leather jacket drives a black muscle car down an empty desert highway at impossible speed. The road stretches endlessly into the glowing sunset. In the rearview mirror, three sleek black drones rise from the heat haze behind her. Camera: low front tracking shot inches above the asphalt, car grille vibrating with engine power.

[0:03–0:06] The drones accelerate, firing thin blue energy bolts that tear sparks across the road. She grips the wheel, jaw tight, then yanks the car into a violent drift around an abandoned gas station. Camera whip-pans from the drones to the car, speed-ramp into slow motion as dust explodes around the tires.

[0:06–0:09] One drone dives low beside the driver window. She grabs a flare gun from the passenger seat, kicks the door open mid-drift, and fires. The flare streaks through dust and sunlight, hitting the drone dead center. It erupts into a spinning fireball above the highway.

[0:09–0:12] The remaining drones split left and right. She drives straight toward a collapsing old steel bridge, engine roaring. The bridge deck begins to break apart beneath the car as she hits a nitro boost. Camera switches to ultra-wide aerial shot: the car launches across the broken gap while debris falls into the canyon below.

[0:12–0:15] Mid-air, she looks calmly at the last drone reflected in her side mirror. The car lands hard, suspension sparking, and the final drone smashes into the falling bridge behind her in a massive explosion. She drives into the sunset as fire blooms in the background — hard cut to black.

Ultra-realistic Hollywood sci-fi chase, desert highway, drones, dust, fire, practical-feeling explosions, anamorphic lens flares, aggressive camera movement, speed ramps, intense motion blur, epic scale, no distortion, no stretching.

Cyber Punk Action

Cinematic surreal action short film, 10s, ultra-realistic, Arri Alexa Mini, 35mm lens. Lighting: Midnight rooftop in a futuristic city, deep blue moonlight mixed with magenta neon signs, wet concrete reflecting skyline lights, drifting mist, strong rim light, shallow DOF, high-contrast cinematic look.

[0:00–0:02] A young courier with silver-dyed hair, reflective jacket, and black gloves sprints across a rain-soaked rooftop. The city glows beneath him, neon signs flickering through the mist. Camera: low-angle tracking shot beside his feet as water splashes with every step. His breathing is calm but intense.

[0:02–0:04] He reaches the edge of the rooftop and leaps across a wide gap between buildings. Time subtly slows mid-air. His jacket catches the wind, raindrops streak past like glowing lines, and the camera rotates around him in a smooth 180-degree arc.

[0:04–0:06] As he lands on the next rooftop, the impact sends a circular pulse of blue light through the puddles. The pulse transforms reflections in the water into animated holographic arrows pointing forward. He doesn’t stop — he slides under a spinning rooftop fan, sparks flying inches above his back.

[0:06–0:08] A small delivery capsule on his wrist opens, projecting a glowing digital butterfly made of light. It flies ahead, weaving through antennas and steam vents. He follows it at full speed as the camera switches to a fast FPV-style chase shot.

[0:08–0:10] He dives through a cloud of rooftop steam and emerges on the edge of a glowing billboard. The butterfly lands on his finger, turns into a bright data spark, and shoots into the skyline. He looks into camera, smirks, then steps backward off the billboard — hard cut to black before he falls.

Ultra-realistic, cinematic cyberpunk atmosphere, dynamic camera movement, rain reflections, neon rim lighting, smooth speed ramps, intense motion blur, no distortion, no stretching.

Was ist Seedance 2.0?

Seedance 2.0 ist das aktuelle KI-Videomodell von ByteDance, dem Konzern hinter TikTok. Veröffentlicht wurde es Anfang 2026 und es zählt seitdem zu den besten Modellen für cinematische KI-Videos auf dem Markt.

Was Seedance 2.0 besonders macht: Es ist ein multimodales Modell. Du kannst nicht nur Text-Prompts eingeben, sondern auch Bilder, Videos und Audio als Referenzen mitliefern – und das Modell verarbeitet alle Eingaben in einem einzigen Generierungsdurchlauf zu einem fertigen Video mit nativem Sound.

Die wichtigsten Eckdaten:

• Videolänge: bis zu 15 Sekunden pro Generierung
• Auflösung: maximal Full-HD (1080p)
• Seitenverhältnisse: 16:9, 9:16, 1:1, 4:3, 3:4, 21:9
• Native Audiogenerierung: Soundeffekte, Musik und Lippensynchronisation werden direkt mitgeneriert
• Multi-Shot in einem Take: Mehrere Szenen mit Schnitten und Übergängen in einem einzigen Output
• Character Consistency: Gesichter, Kleidung und Stile bleiben über mehrere Shots stabil

Vor allem bei komplexer Bewegung, Physik (Wasser, Feuer, Stoffe) und cinematischen Kamerafahrten spielt Seedance 2.0 in der absoluten Top-Liga.

Wie du Seedance 2.0 nutzen kannst

Seedance 2.0 findest du mittlerweile in vielen Multi-KI-Tools, die mehrere Top-Modelle unter einer Oberfläche bündeln. Das ist meistens die praktischere Variante, als das Modell isoliert zu nutzen – du hast in einem Tool gleich Zugriff auf Seedance, Sora 2, Veo 3.1, Nano Banana und Co. und kannst Workflows kombinieren.

Drei Plattformen, die ich besonders empfehlen kann:

• Higgsfield AI (→ direkt zum Tool*) – Multi-KI-Plattform mit Seedance 2.0, Sora 2, Veo, Kling und vielen weiteren Top-Modellen. Mein persönlicher Favorit für KI-Bilder und KI-Videos.
• Magnific – Komplette KI Creative Suite mit Seedance 2.0, Nano Banana, Kling, Sora und ElevenLabs – alles unter einem Dach zu einem fairen Preis.
• MagicFit – Aktuell als Lifetime-Deal auf AppSumo* verfügbar – inklusive aller Seedance-Funktionen (Full-HD, Start-/Endframe, Multi-Referenzen).

Die eigentliche Modell-Qualität ist bei allen Plattformen identisch (es ist dasselbe Modell), Unterschiede gibt es bei den Zusatz-Features, dem Preis-Modell und der Benutzeroberfläche.

Tipps zu Seedance 2.0 Prompts

Damit du das Maximum aus Seedance 2.0 holst, hier die wichtigsten Prompt-Prinzipien:

Strukturiere deinen Prompt klar. Eine bewährte Reihenfolge: Shot-Typ → Subjekt → Aktion → Umgebung → Kamerabewegung → Beleuchtung → Stil → Seitenverhältnis. Diese Struktur reduziert Fehlinterpretationen massiv.

Sei konkret bei der Kamera. „Cinematic shot“ ist zu vage. Schreib stattdessen „low-angle tracking shot“, „handheld FPV“, „dolly-in“ oder „anamorphic 40mm lens“. Seedance versteht Filmsprache erstaunlich gut, also nutze sie.

Definiere die Bewegung. Das Modell ist auf Motion optimiert. Wenn du nicht beschreibst, was sich bewegen soll und wie, wird der Output statisch oder unvorhersehbar.

Beleuchtung und Mood explizit angeben. Begriffe wie „golden hour backlight“, „neon reflections on wet asphalt“ oder „deep shadows, shallow DOF“ verändern das Ergebnis enorm.

Bei längeren Videos: Timeline-Prompting. Teile die 15 Sekunden in Abschnitte auf (z.B. [0:00-0:03], [0:03-0:06]) und beschreibe pro Beat 2-3 Sätze. So bekommst du echte Multi-Shot-Sequenzen statt einer statischen Szene – dieses Muster siehst du auch in meinen Beispielen oben.

Negative Prompts nutzen. „No distortion, no stretching, no warping“ am Ende des Prompts spart oft mehrere Versuche.

Dialoge in Anführungszeichen. Wenn jemand etwas sagen soll, schreibst du: He looks at her and says: "We need to leave. Now.". Seedance generiert die Lippensynchronisation dann direkt mit.

Mit Referenzen arbeiten

Eine der mächtigsten Funktionen von Seedance 2.0 ist die Arbeit mit Referenzen. Du bist nicht auf reine Text-Prompts angewiesen, du kannst dem Modell zeigen, was du willst.

Folgende Referenz-Typen werden unterstützt:

• Startframe: Ein Bild, das als erster Frame deines Videos dient
• Endframe: Ein Bild, das als letzter Frame definiert wird – das Modell generiert den Übergang dazwischen
• Bilder: Für Charakter-Konsistenz, Style-Anker oder als Szenen-Referenz
• Videos: Um Kamerabewegung, Tempo oder Choreografie zu übernehmen
• Audio: Für Beat-Sync, Musikrhythmus oder als Stimmungs-Referenz

In Summe kannst du bis zu 15 Referenzen pro Video kombinieren (9 Bilder + 3 Videos + 3 Audio-Files). Das eröffnet komplett neue Workflows: Du nimmst ein Bild als Startframe, ein zweites Bild für den Hauptcharakter, ein Referenz-Video für eine bestimmte Kamerabewegung und einen Audio-Track für den Rhythmus – alles in einem einzigen Prompt.

Im Prompt referenzierst du sie direkt, z.B. The character from [Image1] performs the dance from [Video1] oder Use [Image1] as the first frame, with the camera movement from [Video1].

MagicFit unterstützt all diese Funktionen – Start- und Endframe, Multi-Image-, Video- und Audio-Referenzen. Aktuell gibt es dafür einen Lifetime-Deal auf AppSumo* (siehe Hinweis oben im Beitrag).