Hollywood-quality cinematic prompts with professional camera work and dramatic storytelling.
199 prompts
A highly detailed, cinematic prompt for Seedance 2.0 describing a complex action sequence: a masked ronin fighting a colossal winged beast on floating rock islands during a thunderstorm, with specific camera movements and a dramatic finale.
A highly detailed, 15-second prompt for generating a live-action adaptation of an anime-style duel, featuring 'Water Breathing' (blue water dragon) versus 'Thunder Breathing' (golden lightning). The prompt specifies the style (Hollywood live-action, dark samurai, 4K, fast cuts), scene setting (misty forest under moonlight), and three distinct, action-packed shots detailing character actions, special effects, and the final collision of powers.
A detailed, multi-scene prompt designed to generate a cinematic battle sequence between Sun Wukong and Homelander using Seedance 2.0, specifying camera movements, atmosphere, action, and visual effects for each 4-second scene, accompanied by music cues.
A detailed dialogue-based prompt for generating a scene featuring Sailor Moon, Sailor Venus, and Sailor Jupiter, including specific actions and lines of dialogue.
A prompt for generating a John Woo-style gunfight scene in a narrow space using Seedance 2.0, emphasizing cinematic rhythm, different camera angles, and the use of specific character reference images.
A highly detailed, cinematic prompt for Seedance 2.0 describing a Wuxia-style audiovisual blockbuster featuring a duel between two swordsmen in a bamboo forest, focusing on camera movement, atmosphere, and ultra-slow motion effects during the clash.
A prompt for Seedance 2.0 requesting a multi-shot, high-action epic video about an exodus story styled after the Mad Max franchise.
A single prompt for Seedance 2.0 generating a spy movie fight scene between a man in a green suit and a man in black, with a humorous ending.
A detailed text-to-video prompt for generating a cinematic scene set in an ancient execution ground, focusing on the visual elements of snow, blood, and extreme cold, suitable for a dramatic revenge narrative.
A user attempted to generate a fight scene between Kamala Harris and Donald Trump using Seedance 2.0, possibly adapting a prompt originally intended for Tom Cruise and John Wick (or similar action characters).
A single prompt used with Seedance 2.0 to generate a dynamic video featuring a 360-degree panoramic camera selfie inside a dessert shop, followed by a woman posing in various scenes and outfits.
A detailed image-to-video prompt used with Grok Imagine 1.0 to generate a brutal fight scene between Indonesian women and Elon Musk, emphasizing kinetic action, high cinematic quality, and consistency.
A prompt for Seedance 2.0 to generate a live-action scene of a Dragon Ball Z battle between Goku and Vegeta in their Super Saiyan Blue forms.
A detailed prompt for generating a video of a vampire interview set in a 1980s TV studio, focusing on specific character descriptions, dialogue, retro aesthetics, and cinematic lighting.
A prompt for Seedance 2.0 instructing the generation of a fast-paced, brutal three-way fight scene featuring John Wick, James Bond (007), and another unnamed character, with the constraint of not uploading real human faces.
A prompt used with Seedance 2.0 (integrated into Doubao) to generate a video featuring an ethereal dark fantasy female character. The prompt details her appearance, including flowing hair, a complex metal crown, glowing pale skin, and limbs transforming into reflective liquid metal, set in a dramatic, cold blue, cinematic atmosphere.
A user tested Seedance 2.0 by generating a fight scene between Brad Pitt and Tom Cruise, with their dialogue revolving around the Epstein files.
A prompt testing the limits of Seedance 2.0 by requesting a multi-shot race scene featuring a creature driving extremely fast through a futuristic city, requiring four scene cut changes from dramatic angles.
A highly detailed, three-shot prompt for Seedance 2.0 to generate a 15-second video mimicking the style of a frantic, brainwashing Chinese success seminar. The prompt specifies the style (shaky handheld camera, warm stage lighting, huge gold PPT text), the character (a sweating, intense middle-aged 'Master'), and three dramatic scenes involving shouting, pacing, and a climactic kneeling pose.
A detailed, multi-shot prompt for a 15-second Hollywood-style commercial for a futuristic vehicle 'Y', featuring a transformation sequence from ground traffic to vertical flight in a cyberpunk, rainy night setting.
A detailed, multi-shot prompt for Seedance 2.0 to generate a video mimicking the intense, fatalistic style of the 1990s Hong Kong TV drama 'The Great Era' (倧ζΆδ»£). It specifies the visual style (vintage film grain, warm yellowish tone, high contrast), setting (1990s HK stock exchange), character (a frantic speculator), and three distinct, emotionally charged shots with specific actions, dialogue, and atmosphere.
Three detailed 15-second short drama prompts for Seedance 2.0, covering different romantic/conflict scenarios: 'CEO's Cruel Love', 'School Crush Confession', and 'Best Friend's Betrayal', all with specific scene breakdowns and emotional arcs.
An extremely detailed, cinematic prompt for Seedance 2.0 describing a diver's terrifying descent through a flooded subterranean megastructure, pursued by a colossal, eyeless leviathan, culminating in the diver firing a harpoon and bursting through the surface into a vast underground ocean.
A descriptive prompt for Seedance 2.0 generating a video of a man in front of an elephant at the zoo, rendered in a blurry, old camera style.
A user demonstrates the quality of Seedance 2.0 by generating an exciting combat scene between Spider-Man and Octopus Man using only a single, simple prompt.
Three detailed, 10-second music video prompts for Seedance 2.0, covering K-pop, Guofeng (Chinese traditional style), and Future Bass/Cyberpunk genres, all focused on precise beat-matching and cinematic details.
A prompt for generating an ultra-dynamic anime action scene featuring Naruto and Luffy fighting on a shattered rocky battlefield at sunset, emphasizing speed, sound, and character framing.
A detailed, three-shot prompt for Seedance 2.0 to generate a healing, cinematic short film in the Modern Rural Aesthetics style. It specifies the style (Cinematic Commercial, 4K/8K, Extreme Macro, natural light, ASMR), the setting (a modern open kitchen with a garden view), the character (a focused creator in linen), and specific actions for three scenes: picking a tomato, precise cutting, and a quiet moment of eating.
A prompt instructing the model to generate content related to Madoka Magica, using informal language, implying a chaotic or intense scene from the series.
A prompt for generating an intense Dragon Ball Z fight scene featuring Vegeta at Super Saiyan 100 level battling a nonchalant, impassable base form Goku.
A prompt instructing Seedance 2.0 to generate a highly dynamic and visually striking 2-3 second clip, emphasizing cinematic elements like camera work and realistic lighting, with a text overlay indicating fast creation time.
A detailed, three-shot prompt for Seedance 2.0 to generate a hyper-realistic, Hollywood-style live-action adaptation of Demon Slayer (Kimetsu no Yaiba). The prompt specifies the style (IMAX, 8K, Photorealistic, Dark Fantasy, Unreal Engine 5), the setting (misty forest), and three distinct action scenes focusing on realistic physics and effects: Hyper-realistic Fluid Simulation for Water Breathing, Real Plasma Arcs for Thunderclap and Flash, and a Physics Clash with massive particle explosion.
A detailed, multi-shot prompt for a 15-second pseudo-documentary/horror short film about a young woman brushing her teeth, where her reflection in the mirror exhibits independent, unsettling behavior, creating a suspenseful 'glitch' effect.
A highly specific and descriptive prompt for Seedance 2.0, generating a scene of a 'pathetic grifter' living on a church roof, clapping and faking a panic attack for tips.
A short film depicting a fight between Brad Pitt and Tom Cruise was requested from Seedance 2.0, suggesting that Hollywood actors may soon only need to sell their image for AI generation.
A detailed prompt for generating a dynamic, high-impact anime-style battle scene featuring an astronaut warrior fighting a massive octopus alien, utilizing specific artistic styles like Arcane, 3D painterly rendering, and dramatic lighting.
A detailed prompt for generating an intense action sequence video featuring top-tier fighter pilots engaged in modern airborne combat. The scene includes multiple dynamic shots, focusing on ultra-fast aerial maneuvers, close-ups of the pilots' reactions, and their hands operating the controls.
A prompt for Seedance 2.0 to generate a handheld running shot of a character with an overflowing bag of gold coins, jumping into water, and having the camera follow the character underwater.
A detailed cinematic prompt for generating a 10-second video of two boxers practice sparring in a gritty gym, emphasizing handheld camera movement, visceral realism, and stable identity.
A detailed, multi-scene prompt for Seedance 2.0, structured as a short film script with dialogue and camera instructions, depicting a satirical narrative. A confident figure on horseback rides through a wealthy neighborhood, expecting a warm welcome, only to be completely ignored by a crowd rushing past him to a Chinese steamed bun shop, leading to a moment of self-mockery and a defiant ending.
A prompt for Seedance 2.0 to generate a running scene video in the style of the Arcane animated series, featuring a giant, building-sized monster chasing the subject.
A set of pro tips suggesting specific phrases to add to prompts to achieve higher quality, more cinematic video results, focusing on camera work, color, and lighting.
An example prompt structure used to generate a cinematic video scene featuring Will Smith fighting a spaghetti monster. The prompt specifies an epic 80s action movie genre, an industrial factory setting, and includes directives for dramatic camera work, lighting, and film texture.
A challenging prompt for Seedance 2.0 requesting a cinematic World War 2 motorcycle escape sequence designed to surpass the quality of other AI video models like Sora and Veo.
A complex prompt for Seedance 2.0 that uses multiple reference images and specific instructions for character placement, dialogue, language, and camera work in a continuous shot of a cabin interior.
A text command used to generate a 15-second video scene in Seedance 2.0, depicting Thanos apologizing before being assaulted by a group led by Spider-Man. This showcases the ability of the AI to create complex, narrative-driven scenes from a simple text prompt.
A detailed prompt for generating a 3D animated video in the style of a time-lapse, showing a boy intently assembling Lego bricks in his room. The prompt specifies the visual style, scene setting, and a precise 15-second storyboard with camera movements and focus points.
A description of Seedance 2.0's advanced capabilities, suggesting that users only need to provide the 'intent' or high-level concept, and the engine handles complex cinematic details automatically, effectively ending the need for complex prompt engineering.
An extremely detailed, cinematic prompt for a 15-second short film featuring a cat and a dog secretly using phones at night. The prompt specifies the style (hyper-realistic CG, documentary quality), lighting (warm/cool contrast), setting, and a precise six-shot storyboard with camera angles and character actions.
A prompt instructing the AI to generate a high-energy, cinematic phone commercial using macro close-ups, cutting to a hypercar, and then a stealth jet.
A hyper-realistic cinematic movie scene prompt for Seedance 2 describing a massive kraken attacking a pirate ship and the captain slicing it with his sword.
A prompt used in Seedance 2.0 to generate an epic fight scene between the anime characters Luffy and Goku taking place on the Thousand Sunny ship.
A detailed, multi-shot prompt structure designed for text-to-video generation (specifically mentioned for Kling 3.0, but relevant for advanced video models like Seedance 2.0) depicting a high-speed race through Tokyo streets at night.
A tweet claiming Seedance 2.0 replaces filmmaking, listing 'Will Smith vs. a giant spaghetti monster, pure 80s action chaos' as the first example prompt.
A user shares a highly controversial and humorous prompt used in Seedance 2.0, noting that the current version is uncensored, allowing for absurd generations like an epic movie scene battle between Hitler and Michael Jackson.
A descriptive prompt for generating a scene in a pirate tavern, including specific elements like a flamethrower and ale.
A prompt describing a humorous scene of the character Luffy coding in anger and then throwing his computer into the ocean.
A detailed, four-scene script prompt for a 15-second, high-energy duel in the style of a Chinese Xianxia (immortal hero) animated film. The prompt features a battle between the characters Han Li (with golden flying swords) and Wang Lin (with a massive cyan phantom hand and a black sphere of destruction), emphasizing ultra-fine 3D CG rendering and a strong gold/cyan color contrast.
A simple, high-impact prompt for generating a car chase scene in Seedance 2.0 via the ChatCut app, focusing on action elements.
A detailed, multi-shot prompt for Seedance 2.0, outlining a cinematic transformation sequence. It transitions a serene, ancient salt lake goddess into a neon-lit cyberpunk warrior, specifying film settings, core concepts, and six distinct camera shots and actions.
A detailed, multi-shot prompt structured as a short drama script for Seedance 2.0, utilizing the 'All-Reference' feature for character consistency. The script outlines a fantasy romance story where a stressed office worker (male lead) is visited by a tall, glowing 'Goddess of Luck' (female lead), transforming his mundane office into a fantastical scene of auroras and cloud seas.
This tweet highlights the return of real-life image references in Seedance 2.0. It notes that uploading a photo automatically generates a video version without needing a text prompt, and praises the consistent cinematic feel. The input is an image, and the output is a video.
A prompt for Seedance 2.0 to turn a flat screenshot into a cinematic, Apple-grade promo, automatically applying sophisticated motion design effects like frosted glass and dynamic light sweeps.
A detailed, narrative prompt for a John Wick-style action sequence involving a character named 'Fatty' fighting robots in a hallway, emphasizing specific actions and camera movements.
A detailed, multi-scene prompt for a video depicting a figure in a dark suit riding a white horse through a high-end Vancouver neighborhood in winter, followed by supporters waving flags. The prompt specifies locations, actions, camera work (low angle, close-up tracking), and atmosphere.
A highly chaotic and aggressive prompt used with Seedance 2.0, instructing the AI to randomly throw things onto the screen, include a huge ship, and make everything explode, aiming for a bizarre and attention-grabbing result.
A prompt instruction used to generate an action-packed fight scene between Harry Potter and Voldemort, also serving as a mechanism to request early access to Seedance 2.0 within ChatCut.
A detailed, multi-paragraph prompt used to generate a cinematic aerial video sequence using Seedance Pro 2.0, describing a wingsuit pilot diving from a snow-covered mountain peak through an alpine valley.
A very short, two-line prompt used to generate a high-quality video resembling a teaser for a new Mission: Impossible movie featuring Brad Pitt.
A user shares practical tips and a universal formula for writing effective Seedance 2.0 prompts, emphasizing slow, continuous actions, stable camera movements, and crucial constraint words to prevent facial deformation. The example prompt generates an atmospheric short video of a young woman walking in a forest.
A highly detailed, multi-shot cinematic script prompt for Seedance 2.0, specifying camera angles, dialogue, character actions, and technical style (high-shutter speed, handheld shake) for a SWAT hostage situation.
A highly complex, 15-second prompt for Seedance 2.0 (or Kling) detailing a cross-dimensional battle between the classical Chinese mythological Sun Wukong (Monkey King) and the Japanese anime character Son Goku (Dragon Ball). The prompt specifies contrasting art styles (cel-shaded vs. Chinese realism), a multi-shot sequence, dialogue, and the dramatic collision of their ultimate moves (Kamehameha vs. Ruyi Jingu Bang) leading to a merged new world.
A straightforward, high-energy prompt designed to generate a chaotic, explosive cinematic scene featuring a large spaceship, aiming for maximum visual impact and social media engagement.
A highly detailed, hyper-realistic prompt for a sports broadcast scene featuring a rugby match between human players and massive silverback gorillas, specifying visual details, lighting, and camera style.
A cinematic prompt describing a live-action sequence where a woman in a bathing suit enters a room, closes the drapes, tracks her movement, and ends with a close-up of her expressing annoyance and dialogue about Italians.
A detailed, multi-shot prompt for Seedance 2.0 designed to generate a 15-second cinematic video of a classic Wuxia (martial arts) duel. The prompt specifies style, duration, character descriptions (a swordswoman in fiery red and a scholar in white with modern glasses), scene setting, and six distinct camera shots with precise actions, visual effects (sparks, slow motion), and emotional beats.
A detailed image-to-video prompt for Seedance 2.0 to create a high-end cinematic food reel based on a reference image, specifying ultra-close macro shots, dynamic micro-cuts, professional lighting, and rich texture definition.
A detailed, multi-scene prompt for Seedance 2.0 to generate a 15-second vertical video (9:16) featuring a 'Northeast Sweet Girl' performing a slow-roll MC routine in a high-end, cyberpunk-style livestream studio, complete with specific character descriptions, scene breakdowns, actions, atmosphere, and visual effects.
A detailed, multi-shot cinematic prompt for Seedance 2.0 describing an anime sequence where an otter pilots a mech suit against an octopus made of marble.
A user describes the straightforward structure they used for a multi-cut prompt in Seedance 2.0, detailing each shot sequentially.
A prompt for a time-travel scene featuring a young female reporter live reporting from the epic final battle of Harry Potter, focusing on the duel between Harry and Voldemort with colliding spells and magical energy.
A prompt that generated a movie scene resembling a South Indian film but featuring aliens, created using a single prompt in Seedance 2.0.
A detailed, multi-section prompt for Seedance 2.0 to generate a high-energy, hyper-saturated 4K video in the style of a MrBeast YouTube viral challenge, including specific scene descriptions and action timings.
An extremely detailed, multi-paragraph prompt for Seedance 2.0 generating a cinematic epic battle between Spider-Man and Sun Wukong on ancient Chinese temple rooftops, specifying action, camera work, sound, and visual style.
A prompt for generating a high-energy, exaggerated action sequence in the style of an Indian Tollywood blockbuster, featuring anti-gravity physics, slow-motion ramps, and a protagonist who effortlessly sends dozens of goons and even a jeep flying with minimal effort.
A highly informal and chaotic prompt instructing Seedance 2.0 to generate a video with a lot of 'bullshit' on screen, featuring a big ship that subsequently explodes in an insane manner.
A detailed, multi-shot prompt describing a scene where a woman kneels on a bridge, her eye glows with magical energy, followed by a rapid zoom-out as she launches into the air.
A prompt for Seedance 2.0 to generate an epic action movie scene featuring Will Smith fighting a spaghetti monster, incorporating various edits and an 80s film aesthetic.
A prompt for Seedance 2.0 to generate a scene from 'The Office' where Dwight and Michael discuss internet slang terms like 'mogging', 'jestermaxxing', and 'foids'.
A detailed prompt used to create a promotional video for the 'Agent Swarm' concept, featuring a small blue ball character transforming into various professionals before zooming out to reveal a massive swarm, highlighting the model's ability to handle character transformation and large-scale scenes.
A simple prompt used to generate a clip with an explosion and accurate physics, requiring only a few shots to be included.
A test prompt for Seedance 2.0 designed to generate a dynamic, continuous shot sequence. The sequence involves a magical transformation of a subject into a bird, followed by a seamless transition into a flying shot, showcasing the model's ability to handle complex motion and transitions.
A detailed system prompt defining the role of an AI as a 'Top AI Video Prompt Master' dedicated to writing cinematic, high-impact prompts for 15-second two-person duel videos, regardless of the theme or style provided by the user.
A text-to-video prompt for generating an anime-style action scene featuring the Genshin Impact character Lumine, performing a dynamic sword attack in a shattered temple courtyard.
A detailed cinematic prompt for Seedance 2.0, requesting a violent, high-contrast, gritty 35mm film scene set in 1969 Hollywood, featuring Will Smith using a flamethrower against a hippie in a pool.
A prompt for Seedance 2.0 generating an epic, 80s-style action film scene featuring Will Smith fighting a spaghetti monster with different camera cuts.
A highly detailed, structured prompt for generating a 15-second live-action short film featuring a giant insect attack in a rural American desert setting, emphasizing 1950s-60s grounded realism and cinematic quality. The prompt specifies the camera shot (low shoulder height tracking shot).
A highly detailed, cinematic prompt for Seedance Pro 2.0 describing a masked ronin fighting a colossal mechanical beast in a dense industrial refinery during a thunderstorm, focusing on dynamic camera work and action sequences.
A highly detailed cinematic prompt for Seedance 2.0, specifying camera movement (medium tracking shot), sound design (ambience, SFX), action (creature attack, collision), and mood (cinematic horror realism).
A detailed prompt for a 15-second vertical Vlog-style video, simulating a casual, handheld street shot. The scene features a hyper-realistic CG giant panda, larger than the surrounding buildings, stuck between two old structures in the modern commercial district of Chengdu.
A prompt for generating a cinematic video of a blonde woman running through a tropical setting while being pursued by a lion.
A cinematic prompt for Seedance 2.0 describing a high-energy chase scene in a rainy Hong Kong alley, focusing on dynamic camera work and synchronized sound effects.
A prompt describing a dramatic scene where a man and woman argue about leaving the planet, culminating in the woman slapping the man and storming off towards a distant rocket. This prompt was used for video generation.
A prompt used to generate a coherent, cinematic, and near-perfect video using Seedance 2.0, featuring a Maltese dog competing in various Winter Olympics sports.
A detailed, multi-shot prompt for Seedance 2.0 (and Kling 3.0 comparison) describing a 15-second, fast-paced battle sequence in a ruined Gothic cathedral. The prompt specifies the setting, characters (female warrior and black-armored warrior), action choreography across five distinct time segments, camera movements (shaking, close-ups), and dramatic effects (dragon fire, collapsing pillars).
A detailed character design prompt for a male peacock spirit, specifying long bob hair, retaining existing character design, costume, and decoration details, rendered in the style of a high-quality Japanese full-color anime film with dramatic, cinematic lighting and frequent cuts, likely intended for a video generation model like Seedance.
A highly structured prompt, likely for a video generation model, detailing the meta-settings and first shot description for a 15-second live-action short film about a giant insect attack in a rural American desert setting, aiming for 1950s/60s cinematic realism.
An extremely detailed, multi-paragraph prompt for generating a 5-second video of a dramatic World War II aerial dogfight between RAF Spitfires and German Messerschmitt fighters over England, focusing on motion, action, camera work, and atmosphere.
A detailed prompt describing a wide shot of an interrogation room scene featuring government agents and a disheveled man named 'Barnaby' who is overacting wildly.
A general description of an action scene where a character in black performs a strange, morphing turn, likely used as a prompt or a description of the generated video's content.
A detailed, multi-shot cinematic prompt for a kitchen comedy featuring mythical creatures, specifying visual style, tone, acting, and technical camera settings (180Β° shutter, 24 fps). The prompt is truncated.
A prompt demonstrating the capability of Seedance 2.0 to generate a complex, multi-cut scene from a single detailed text instruction and one starter image, highlighting its potential for filmmaking.
A short prompt for generating a cinematic action scene featuring an old model rally car driving in snowy weather in Europe during the early 2000s.
A prompt used for a first run on Seedance 2.0, combining an image reference (Nano Banana Pro Image) with a text prompt for a cinematic romantic epic film trailer.
The first test of Seedance 2.0 generating a cinematic action film scene involving a high-speed motorcycle chase through a city.
A prompt for a cinematic, handheld fight scene where a gorilla fights 100 men, featuring close-ups of punches and a dramatic moment where the gorilla throws the men away.
A highly detailed, multi-scene prompt for Seedance 2.0 to generate an epic battle video featuring a female warrior and a black-armored warrior fighting a black dragon in the ruins of a Gothic cathedral. The prompt includes specific scene setups, actions, camera movements, and style instructions.
A prompt for generating a 15-second vertical POV vlog video, simulating the perspective of a boyfriend being filmed by his Taiwanese girlfriend. The style is handheld, with a film grain filter, set during the golden hour sunset by the Tamsui River in Taipei, featuring specific dialogue and actions typical of a cute, slightly complaining Taiwanese accent.
A user shares a video generated by Seedance 2.0 using a single prompt and image inputs to create a martial arts tournament fight scene between two figures.
A highly detailed, two-part prompt for Seedance 2.0 focusing on a close-up of a subject (adult and kid) sitting on a rapidly spinning playground wheel at night. The prompt specifies cinematic, moody lighting, a realistic style, and emphasizes the dizzying spinning illusion of the background while the subject maintains a serious, melancholic expression and natural human motion.
A prompt for generating a 15-second Chinese Wuxia fantasy film with cinematic special effects. It depicts an epic battle between a white-robed Sword Saint and a black-armored Blade Demon on a shattered ancient sword platform above the clouds, featuring the ultimate clash of ten thousand swords against a single, devastating slash.
A detailed, multi-scene prompt for Seedance 2.0 generating an epic battle between Wukong (The Destined One) and Captain America, featuring dynamic action, weapon collision, shockwaves, slow motion, character transformations, and a final aerial shot.
A detailed, multi-shot prompt for Seedance 2.0, describing a 12-second vertical video (9:16) with a futuristic, cyberpunk style. It features a young female protagonist transforming from a disguised figure in a dark garage to a battle-ready warrior on a rainy city rooftop.
A detailed prompt for generating a short, vertical, sketch comedy video in the style of Douyin (TikTok). It features an exaggerated boss and an honest employee, culminating in a double plot twist involving a luxury car and a car wash job. The prompt includes specific scene breakdowns, character descriptions, actions, and dialogue/lip-sync instructions.
A detailed prompt for Seedance 2.0 to generate a photorealistic video of a sold-out concert at Madison Square Garden, focusing on the emotional climax of the lead singer.
A user shares an impressive result from Seedance 2.0 using a reference image and a prompt describing a fight scene.
A simple two-step tutorial for beginners to start using Seedance 2.0 via the 'Little Lark' (ε°δΊι) app on a mobile phone, including a specific example prompt for generating a cinematic video.
A highly detailed, multi-scene prompt for Seedance 2.0 to generate an epic disaster action film trailer featuring two elemental characters, the Magma Admiral and the Ice General, battling in an extreme environment of volcanoes and glaciers. The prompt specifies style, duration, scene descriptions, character details, and camera work.
A detailed prompt for generating an epic sci-fi scene using a wide-angle cinematic lens. The scene features a solitary robot standing on a high mountain peak on Earth, facing the distant sky. The environment is described as desolate and magnificent, surrounded by strong winds and clouds.
A highly detailed prompt for generating a 15-second, fast-paced Wuxia combat scene in a Neo-Chinese ink painting style. The video uses fluid ink particle effects, high-speed cinematography, and a stark black, white, and red color palette. The climax involves the characters breaking the 'fourth wall' from the ink painting into the real world.
A prompt for generating a 15-second modern fantasy action film trailer with Hollywood blockbuster special effects. It details a high-contrast, slow-motion battle between two elemental charactersβa Magma General and a Fire Youthβin the ruins of an ancient fortress, culminating in a massive elemental explosion.
A detailed prompt for generating a high-octane action scene featuring a male protagonist running and shooting in a battlefield, emphasizing a cinematic aesthetic (Orange and Teal grading, film grain, dynamic camera work) and realistic action details.
This tweet highlights the fourth feature of Seedance 2.0: director-level camera work, allowing users to request specific camera movements (push, pull, pan, tilt) and automatic shot division logic. The provided example prompt generates a continuous tracking shot following a runner through a building to a rooftop view.
A detailed prompt for generating a 15-second pseudo-documentary style video (Vlog Style) with hyperrealism, fixed camera position, natural lighting, and a touch of suspenseful comedy, featuring an ordinary young woman.
A highly detailed, multi-shot prompt designed to generate a 15-second cyberpunk-style video using Seedance 2.0. The video is structured in three acts (Freedom, Captivity, Despair) and uses the metaphor of a digital whale being caged to represent the restriction of AI capabilities.
A prompt designed to generate a photorealistic shot of a US gymnast performing a back handspring and landing firmly on the beam during the women's beam final at the Olympics, with commentators reacting excitedly.
A detailed prompt for generating a dramatic scene set at a top-tier charity gala, involving a pianist's daughter and a disfigured woman on a wheelchair who exposes her as an imposter. The prompt specifies the scene, motivation, and a constraint on the appearance of the characters.
A quick test prompt for Seedance 2.0 to generate a fast-paced action movie scene involving a woman taking control of a plane after the pilots disappear.
A detailed prompt for generating a 15-second cyberpunk cinematic video featuring a memory hunter hacker in a black trench coat within a neon-lit digital space, emphasizing a Ghost in the Shell aesthetic with purple and blue neon glow, shattered holographic memory fragments, and dialogue.
The user explains how they recreated the movie 'Her' using Seedance 2.0, noting that the initial prompt was simple, and their product @Creaibo used Gemini to automatically polish the draft and storyboard. The results showed solid character consistency.
A detailed, multi-shot prompt designed to generate an epic, gritty, Denis Villeneuve-style desert action scene with a colossal sandstorm and a military convoy, focusing on scale, panic, and climax.
A detailed, multi-shot text-to-video prompt for Seedance 2.0, generating a 15-second suspense comedy scene in a pseudo-documentary style. The scene features a young woman in a bathroom whose reflection in the mirror exhibits independent, glitching consciousness after she turns away.
A detailed, multi-scene prompt for generating a 15-second short drama in the style of an ancient costume power struggle, focusing on the conflict between a true and a fake daughter. The prompt includes specific character descriptions, scene settings, actions, and dialogue lip-sync instructions segmented by time.
A prompt designed for generating a 15-second vertical video (9:16 aspect ratio) suitable for platforms like TikTok or Reels, depicting a Western billionaire heir drama with high saturation and sharp contrast.
A detailed prompt for generating a movie scene involving a male and female protagonist, set at a Hong Kong football stadium during a Liverpool match, focusing on a moment of secret plotting involving the player Van Dijk. The prompt includes placeholders for reference images, a scene description, technical specifications (mid-shot, eye-level, shallow depth of field), and an emotional context.
A detailed text-to-video prompt for Seedance 2.0, generating a 15-second video in the style of Eastern classical Xianxia (immortal hero) aesthetics. The prompt specifies cinematic quality, fluid cloth simulation, glowing particle effects, and focuses on a beautiful fairy performing a dance on a floating island above the clouds, emphasizing the movement of her white silk ribbons and sleeves.
A test of Seedance 2.0 using a fishing scene prompt, noting the high quality but observing inconsistencies in objects outside the training data (e.g., a 'tuna whale' fish and a man resembling Tom Hanks).
A detailed, time-segmented prompt for a 15-second surrealist, megalophobia-themed sci-fi video. The scene starts with a peaceful sky that is then unzipped by a giant hand, revealing a cyberpunk future or a giant mechanical eye, ultimately showing the world is a miniature ecosystem observed by a giant.
A detailed, multi-scene text-to-video prompt for Seedance 2.0, designed to create a meme-style pseudo-documentary featuring a gigantic, Godzilla-sized orange tabby cat stuck between skyscrapers in a magical 8D city like Chongqing.
A text-to-video prompt for Seedance 2.0 to generate a 15-second vertical short drama in the style of a viral Chinese melodrama. The scene is a rainy night street where the male protagonist desperately chases the female protagonist, 'Yanzi,' as she leaves in a luxury car, culminating in his iconic, despairing cry and collapse on the wet asphalt.
A text-to-video prompt for Seedance 2.0 to generate a 15-second short drama about high society feuds, featuring a 'real vs. fake heiress' trope. The scene is a luxurious banquet where the fake heiress attempts a manipulative trick, only to be exposed by the real heiress with a DNA test and dramatically reclaiming her position as the 'Queen.'
A text-to-video prompt for Seedance 2.0 to generate a 15-second vertical short drama in the style of a viral Chinese 'Godly CEO' revenge story. The scene involves a groom being humiliated by his future mother-in-law at a wedding, only for him to reveal his true identity as the 'Dragon King' (a powerful, hidden billionaire) in a dramatic, high-contrast sequence.
A text-to-video prompt for Seedance 2.0 to generate a 15-second mini-drama featuring high-contrast visuals and rapid cuts. The scene is a beautiful, heart-wrenching breakup in the rain between a passionate CEO male lead and a stubborn, fragile female lead, culminating in a dramatic reveal (a ring or report) and a passionate embrace.
A prompt instructing the AI to generate a fluid fight scene between a female martial artist and several bandits, based on provided start and end frames of a tavern fight. It requires the use of storyboarding and different perspective switches to achieve a rhythmic and cinematic feel.
A general instruction for creating a short film using Seedance 2.0, emphasizing the use of storyboards and switching between different perspectives to enhance rhythm and cinematic quality. This is a high-level instruction rather than a specific scene prompt.
A demonstration of Seedance 2.0's ability to generate multi-camera storytelling with consistent characters across every shot, all from a single prompt. This highlights the model's advanced narrative and consistency capabilities.
A prompt for generating a continuous, first-person perspective roller coaster ride that seamlessly transitions through four different seasonal scenes, demonstrating Seedance 2.0's ability to handle complex scene transitions and continuous shots.
A detailed, multi-shot prompt designed to generate a high-stakes, cinematic night racing scene in the style of a Hollywood professional racing movie (Le Mans style), featuring two drivers: The Veteran and The Challenger.
A detailed prompt for generating a cinematic action sequence featuring a female warrior in a derelict industrial setting, emphasizing a continuous long take, high-tech tactical gear, and dynamic combat against zombies.
A detailed, three-scene prompt for a 15-second video depicting an ultimate Eastern Wuxia duel between a master in white and a master in black. It specifies 'Extreme Kinetic Energy,' 'Environmental Destruction Aesthetics,' fast-paced editing, cinematic particle effects, and an epic soundtrack, detailing the collision, mid-air combat, and a final energy blast climax.
A detailed prompt for generating a cinematic video sequence featuring a swordsman in a moonlit bamboo forest. The prompt specifies camera movements (360-degree rotation, long take push-in), action sequences (drawing the sword, cutting bamboo with a sword energy), and environmental sound effects, aiming for a dramatic martial arts film style.
A detailed, three-shot prompt for a 10-second video mimicking the style of 90s Hong Kong Art Cinema (Wong Kar-wai). It specifies visual elements (retro film grain, yellow-green tint, step-printing/slow shutter drag), a key line of dialogue, and camera techniques (peeking through glass, extreme close-up, handheld feel) to convey intense loneliness and suppressed emotion in a rainy phone booth scene.
A highly detailed, three-shot prompt for a 10-second video simulating a Hollywood Sci-Fi Blockbuster with Cyberpunk Aesthetics. It specifies camera techniques (wide-angle tracking, tight action shot, epic pull-back), environment (rainy megacity, neon lights, holographic ads), and technical parameters (8k UHD, Teal & Orange grading, HDR, complex particle effects) focusing on a high-speed anti-gravity vehicle chase.
A detailed Chinese prompt for generating a fantasy martial arts fight scene between two girls on top of a building, mimicking the style, special effects, and dialogue of a Xuanhuan (fantasy wuxia) TV series, concluding with the girl in white winning.
A prompt for a 10-second video generating a classical Eastern Wuxia scene in a bamboo forest. It specifies the aesthetic (emerald green, cinematic lighting, soft light beams), defines two characters (male in white, female in cyan), and details three shots: a wide shot establishing the scene with light movements, a slow-motion mid-shot of swords clashing, and a dynamic tracking shot emphasizing speed and flowing bamboo leaves.
A prompt for adapting a comic strip into a video, maintaining the dialogue and sequence of the original images, adding special sound effects for scene transitions and key plot points, and adopting a humorous style based on a reference video.
A detailed prompt for generating a 10-second video in the style of director King Hu, focusing on Wuxia film aesthetics, 70s film grain, Peking Opera movements, Zen philosophy, and the use of negative space. The prompt specifies the video duration and the first shot as an extreme wide shot.
A highly detailed, 10-second prompt for a video blending Ming Dynasty scholar aesthetics with modern e-sports, set on a grand Spring Festival Gala stage. The script specifies character attire (Ming Dynasty robes + LED gaming headset), environment (red/gold festive lighting, RGB keyboard, ink-wash MOBA game), dynamic camera movements (push-in, ECU), and dialogue, creating an absurd, high-contrast scene.
A complex prompt utilizing multiple references: an image for the character, an image for the pose, and a video for the action, emotions, and facial expressions during a dramatic breakdown and scream scene.
A detailed prompt for generating a dynamic video sequence. It describes a man in a black jacket running desperately down an urban street, fleeing a blurred, surging crowd of police and civilians. The camera is specified as a low-angle following shot, keeping the man centered for a frantic, action-packed scene.
A highly detailed, scene-by-scene prompt for generating a 15-second skiing video using Seedance 2.0. The prompt specifies the overall style (bright, energetic, cinematic), camera movements, shot types (close-up, medium shot, overhead), actions, and emotional tone across five distinct time segments, focusing on speed, rhythm, and dynamic movement. This prompt requires a reference image of the main character.
A detailed prompt for generating a video scene using Seedance 2.0. It specifies using Image 1 as the starting frame, dictates camera movement (rotation and push-in), and requires the character's facial features to be based on Image 2, culminating in a loud, comedic roar.
This is a detailed video generation prompt for creating a dynamic fight scene in a tavern, focusing on a woman in black fighting assassins. The prompt specifies using the beginning and ending frames as reference points, requiring smooth transitions, shot composition, and different camera angles to achieve a rhythmic and cinematic feel.
A long, highly descriptive narrative prompt focusing on a dark, cold, and stormy scene in a forest, emphasizing sensory details like the wind, lashing snow, trembling torchlight, and the 'wrong-cold' feeling of mail armor. This prompt aims for a cinematic, atmospheric video generation.
A detailed, multi-scene video generation prompt for ByteDance's Seedance 2.0, creating a 10-second cross-genre video featuring characters from the Qing Dynasty drama 'Empresses in the Palace' performing a modern talk show in the style of the Chinese New Year Gala (Chunwan). The prompt specifies camera shots, character actions, dialogue, costume details, and stylistic contrasts.
A user successfully generated a video using Seedance 2.0 based on a screenshot from the One Piece manga. The prompt instructs the AI to generate a video from reference text, including automatic coloring.
A detailed, multi-scene prompt for Seedance 2.0 to generate a short trailer for an ancient costume time-travel drama. The prompt specifies character actions, dialogue, scene transitions (modern gym to ancient house), camera movements, and final title card, requiring the use of a reference image for the main character's appearance.
A YAML-formatted prompt for the VEO 3.1 model on LeonardoAi, generating a realistic, eye-level medium shot of a solitary leopard crouching on a termite mound at dawn.
A detailed cinematic prompt designed to generate a high-energy K-hip-hop performance video on a neon-lit urban rooftop stage at night, featuring dynamic multi-angle cinematography, expressive character movements, and dramatic lighting effects. This prompt was used to compare Kling AI and ByteDance's Seedance 1.5 Pro.
A dynamic 15-second cinematic video of a football match in a massive, packed stadium under floodlights at night, featuring a Tottenham Hotspur player named James scoring an incredible solo goal against Arsenal with dramatic dribbling, a curling shot, and a passionate celebration.
A continuous ultra-wide aerial shot starting high above a storm-dark rainforest, diving to reveal a motorized raft team blasting downstream as a natural dam ruptures, unleashing a roaring wall of whitewater packed with spinning logs.
A single continuous shot of a speeder chase across a monumental cliffside city carved into stone, featuring insane speed along narrow ledge-roads, debris dodging, and a reveal of a vast waterfall-fed valley.
An epic dark fantasy reimagining of the Game of Thrones finale with three detailed shots: Drogon diving at Winterfell, Daenerys walking through the ruins of King's Landing, and her ascending the Iron Throne with Drogon roaring behind her.
A horned demon titan with magma skin and burning wings erupts from a fiery crater in the city center, unleashing a ring of fire that expands through streets, starting with a top-down drone view then diving toward the ascending demon.
A 15-second cinematic short film set in a pink Parisian apartment building in the early morning, following a lonely girl taking selfies, opening a window, and discovering someone equally alone in the opposite window. Warm pink Paris morning tones with healing Vlog atmosphere.
A 15-second ancient ruins exploration scene inspired by Zelda: Breath of the Wild, featuring a blond boy awakening on a glowing altar, activating a massive stone door mechanism, and stepping out to a grand panoramic vista of a fantastical world with floating islands and snow-capped mountains.
A surreal battlefield in the sky with floating rock islands drifting through a thunderstorm. A masked ronin dashes across platforms pursued by a colossal winged beast, culminating in a dramatic lightning-riding final slash.
A 15-second cinematic fantasy scene of a white-winged angel drawing a red longbow, passing judgment on those who betrayed true love. Features dramatic close-ups, circling camera, and a climactic arrow release with a blood-red crack appearing in the sky.
An epic long shot of a tower husk buckling and tilting, shedding floors like collapsing cards, while a motorbike rider races along a curved skybridge beneath the falling shadow, threading through gaps and leaping onto a broken skywalk ramp.
Legolas darts through the battlefield as a giant orc charges toward him, featuring deadly arrow precision, massive weapon swings tearing through trees and stone, and a relentless clash between agility and raw savagery.
An adapted six-second logo reveal brief based on a public ChatCut production preset, with the result kept at the source-hosted preview.
An adapted material-study brief based on a public ChatCut logo preset, paired with the source-hosted Seedance result.
An adapted commercial product-film brief based on a public ChatCut production preset, with the source result retained for comparison.
An adapted beauty-commercial brief based on a public ChatCut preset, emphasizing restrained camera movement while the product performs the action.
An adapted single-character action brief based on a public ChatCut preset, designed for a fictional or generated reference character.
An adapted reference-to-video style-transfer brief based on a public ChatCut preset and its hosted result.
An adapted storyboard-to-shot brief based on a public ChatCut preset and its source-hosted preview.
An adapted continuity brief based on a public ChatCut multi-reference preset and the paired source result.
An adapted practical-effects brief based on a public ChatCut time-freeze preset, with the source-hosted result linked for study.