[{"data":1,"prerenderedAt":999},["ShallowReactive",2],{"\u002Fblog\u002Fseedance-2-5-prompt-guide":3},{"id":4,"title":5,"authors":6,"body":10,"category":987,"date":988,"description":989,"extension":990,"image":991,"keywords":992,"meta":993,"navigation":126,"path":994,"readTime":987,"seo":995,"sitemap":996,"stem":997,"__hash__":998},"blog\u002Fblog\u002Fseedance-2-5-prompt-guide.md","Seedance 2.5 Prompts Guide: 25 Seedance Prompt Examples",[7],{"name":8,"avatar":9},"Nora Bennett","https:\u002F\u002Fcdn.seedance3.tools\u002Fseedance3-tools\u002Fweb\u002Fassets\u002Fblog\u002Fauthor\u002FNora-Bennett.webp",{"type":11,"value":12,"toc":916},"minimark",[13,17,20,30,33,38,91,95,103,106,109,120,123,137,151,154,157,164,171,176,179,183,186,190,193,197,200,204,207,211,214,218,221,225,228,232,235,244,247,250,256,263,266,274,277,280,284,288,294,300,306,312,320,324,329,335,339,344,350,354,359,365,369,373,378,384,388,393,399,403,408,414,418,423,429,433,437,442,448,452,457,463,467,472,478,482,486,491,497,501,506,512,516,521,527,531,535,540,546,550,555,561,565,570,576,580,584,589,595,599,604,610,614,619,625,629,633,638,644,648,653,659,663,668,674,678,682,687,693,697,702,708,717,720,724,727,731,734,738,741,745,748,752,755,759,762,765,768,771,786,789,795,798,803,809,814,820,828,831,835,838,842,845,849,852,856,859,863,866,870,873,877,891,894,897,900,909],[14,15,16],"p",{},"A beautiful opening shot is not enough if the character changes clothes at second twelve, the camera forgets where it is going, or the dialogue arrives after the speaker stops moving.",[14,18,19],{},"Seedance 2.5 prompts need to direct a sequence, not merely describe a picture. This guide gives you a practical prompt structure and 25 Seedance prompt examples for cinematic scenes, product ads, character consistency, reference-driven motion, dialogue, editing, and video extension.",[21,22,23],"blockquote",{},[14,24,25,29],{},[26,27,28],"strong",{},"Quick answer:"," A strong Seedance 2.5 prompt defines the task, assigns every reference a role, protects the details that must remain consistent, divides longer scenes into clear beats, and coordinates camera movement, action, and sound.",[14,31,32],{},"This guide follows the official BytePlus Seedance 2.5 prompting documentation and provides adaptable templates for its documented generation, reference, editing, and extension workflows.",[14,34,35],{},[26,36,37],{},"In this guide:",[39,40,41,49,55,61,67,73,79,85],"ul",{},[42,43,44],"li",{},[45,46,48],"a",{"href":47},"#what-makes-a-seedance-25-prompt-different","What Makes a Seedance 2.5 Prompt Different?",[42,50,51],{},[45,52,54],{"href":53},"#the-seedance-25-prompt-formula","The Seedance 2.5 Prompt Formula",[42,56,57],{},[45,58,60],{"href":59},"#how-to-assign-image-video-and-audio-references","How to Assign Image, Video, and Audio References",[42,62,63],{},[45,64,66],{"href":65},"#_25-best-seedance-prompt-examples","25 Best Seedance Prompt Examples",[42,68,69],{},[45,70,72],{"href":71},"#common-seedance-25-prompt-mistakes","Common Seedance 2.5 Prompt Mistakes",[42,74,75],{},[45,76,78],{"href":77},"#how-to-edit-specific-details-in-a-seedance-25-video","How to Edit Specific Details in a Seedance 2.5 Video",[42,80,81],{},[45,82,84],{"href":83},"#seedance-25-prompt-faq","Seedance 2.5 Prompt FAQ",[42,86,87],{},[45,88,90],{"href":89},"#create-better-seedance-videos","Create Better Seedance Videos",[92,93,48],"h2",{"id":94},"what-makes-a-seedance-25-prompt-different",[14,96,97,98,102],{},"The official ",[45,99,101],{"href":100},"https:\u002F\u002Fdocs.byteplus.com\u002Fen\u002Fdocs\u002FModelArk\u002F2607689","BytePlus ModelArk Seedance 2.5 prompt guide"," says the model can create a single video up to 30 seconds long and accept up to 50 image, video, and audio reference assets in one request. It also documents subject, motion, style, audio, storyboard, keyframe, editing, extension, and transition workflows.",[14,104,105],{},"That does not mean every prompt should use 30 seconds or 50 files. It means a Seedance prompt can now behave more like a compact production brief.",[14,107,108],{},"For a simple five-second shot, one paragraph may still be enough. For a longer scene, use a timeline:",[110,111,117],"pre",{"className":112,"code":114,"language":115,"meta":116},[113],"language-text","0-5s: Establish the subject, location, and visual identity.\n5-14s: Begin the main action with one clear camera movement.\n14-24s: Develop the action or introduce one controlled change.\n24-30s: Resolve the scene and hold on the intended final frame.\n","text","",[118,119,114],"code",{"__ignoreMap":116},[14,121,122],{},"The time ranges should describe a coherent progression. Do not force four unrelated scenes into one generation simply because more seconds are available.",[124,125,130,131,136],"video",{"controls":126,"playsInline":126,"preload":127,"width":128,"ariaLabel":129},true,"metadata","100%","Official Seedance 2.5 thirty-second storyboard and multi-shot prompt example","\n  ",[132,133],"source",{"src":134,"type":135},"https:\u002F\u002Fcdn.seedance3.tools\u002Fseedance3-tools\u002Fweb\u002Fassets\u002Fblog\u002Fseedance-2-5-prompt-guide\u002Fseedance-2-5-official-30-second-storyboard.mp4","video\u002Fmp4","\n  Your browser does not support the video element.\n",[14,138,139],{},[140,141,142,143,150],"em",{},"Official 30-second Seedance 2.5 storyboard example. Source: ",[45,144,149],{"href":100,"target":145,"rel":146},"_blank",[147,148],"noopener","noreferrer","BytePlus ModelArk",".",[92,152,54],{"id":153},"the-seedance-25-prompt-formula",[14,155,156],{},"Use this formula as a checklist rather than a rigid sentence template:",[21,158,159],{},[14,160,161],{},[26,162,163],{},"Task + Reference Roles + Subject Locks + Timeline + Camera Plan + Visual Direction + Audio + Constraints + Output Goal",[14,165,166],{},[167,168],"img",{"alt":169,"src":170},"Seedance 2.5 prompt formula covering task, references, subject, timeline, camera, style, audio, constraints, and output","https:\u002F\u002Fcdn.seedance3.tools\u002Fseedance3-tools\u002Fweb\u002Fassets\u002Fblog\u002Fseedance-2-5-prompt-guide\u002Fseedance-2-5-prompt-formula.webp",[172,173,175],"h3",{"id":174},"_1-task","1. Task",[14,177,178],{},"Start by saying what Seedance should do: generate a scene, animate an image, follow motion, edit a video, extend a clip, or connect two videos.",[172,180,182],{"id":181},"_2-reference-roles","2. Reference Roles",[14,184,185],{},"Explain what information to take from each file. Use an image for identity, a video for movement, or audio for voice instead of asking every reference to control everything.",[172,187,189],{"id":188},"_3-subject-locks","3. Subject Locks",[14,191,192],{},"Name the details that cannot change: face, clothing, product geometry, label, logo, color, proportions, or voice.",[172,194,196],{"id":195},"_4-timeline","4. Timeline",[14,198,199],{},"For longer videos, write a beginning, development, and ending. Give each beat one main action.",[172,201,203],{"id":202},"_5-camera-plan","5. Camera Plan",[14,205,206],{},"Connect camera movement to the action. One intentional move per beat is usually clearer than stacking orbit, zoom, pan, and handheld movement together.",[172,208,210],{"id":209},"_6-visual-direction","6. Visual Direction",[14,212,213],{},"Define the environment, lighting, texture, and one coherent style.",[172,215,217],{"id":216},"_7-audio","7. Audio",[14,219,220],{},"State the dialogue, ambience, music, or sound effects and when they should happen.",[172,222,224],{"id":223},"_8-constraints","8. Constraints",[14,226,227],{},"Say what must remain unchanged and what should not appear. Constraints are especially useful for product fidelity, character identity, text, and video editing.",[172,229,231],{"id":230},"_9-output-goal","9. Output Goal",[14,233,234],{},"Finish with the intended use, composition, aspect ratio, or final frame when those settings are available in your interface.",[14,236,237,238,243],{},"The same fundamentals - clear action, camera direction, visual style, and constraints - also apply to the earlier model. See our ",[45,239,242],{"href":240,"target":145,"rel":241},"https:\u002F\u002Fseedance3.tools\u002Fblog\u002Fseedance-2-0-prompt-guide",[147,148],"Seedance 2.0 prompt guide"," for shorter prompt structures.",[92,245,60],{"id":246},"how-to-assign-image-video-and-audio-references",[14,248,249],{},"Use direct instructions such as:",[110,251,254],{"className":252,"code":253,"language":115,"meta":116},[113],"Use @Image1 for the woman's face, hairstyle, and green jacket.\nUse @Image2 for the apartment layout and warm window light.\nFollow only the body movement and timing from @Video1.\nUse @Audio1 for the dialogue, voice, and lip movement.\nKeep the character identity and clothing unchanged throughout.\n",[118,255,253],{"__ignoreMap":116},[14,257,258,259,262],{},"The word ",[26,260,261],{},"only"," is useful when a reference has one limited purpose. If you want the motion from a video but not its actor, clothing, or location, say so.",[14,264,265],{},"Use the fewest references that can protect the result. Adding more files can introduce conflicting identities, styles, lighting, and camera logic.",[21,267,268],{},[14,269,270,273],{},[26,271,272],{},"Reference mode vs. editing mode:"," In a reference task, uploaded media guides a newly generated video. In an editing task, Seedance modifies the uploaded video itself, so the source duration or aspect ratio may remain fixed. Check the selected mode before writing your prompt.",[92,275,66],{"id":276},"_25-best-seedance-prompt-examples",[14,278,279],{},"These Seedance prompt examples are original templates. Replace the bracketed details, confirm that the required mode is available, and test one important variable at a time.",[172,281,283],{"id":282},"text-to-video-and-story-prompts","Text-to-Video and Story Prompts",[172,285,287],{"id":286},"_1-cinematic-coffee-shop-one-take","1. Cinematic Coffee Shop One-Take",[14,289,290,293],{},[26,291,292],{},"Use case:"," Atmosphere, natural action, and native sound.",[110,295,298],{"className":296,"code":297,"language":115,"meta":116},[113],"Generate a cinematic one-take scene inside a quiet specialty coffee shop at sunrise. 0-5s: begin on a macro close-up of coffee beans falling into a grinder. 5-14s: track sideways as a barista in a forest-green apron grinds the beans and prepares an espresso. 14-24s: follow the cup to a wooden table beside the window as steam rises into the warm light. 24-30s: slowly push in on the finished coffee and hold. Warm 35mm film look, shallow depth of field, realistic hand movement. Include grinder sounds, soft cafe ambience, and distant morning traffic. Keep the barista's appearance and apron consistent. No text or logos.\n",[118,299,297],{"__ignoreMap":116},[14,301,302,305],{},[26,303,304],{},"Why it works:"," The action, camera, and sound advance together toward one clear final frame.",[124,307,130,309,136],{"controls":126,"playsInline":126,"preload":127,"width":128,"ariaLabel":308},"Official Seedance 2.5 timed prompt example featuring a panda cub",[132,310],{"src":311,"type":135},"https:\u002F\u002Fcdn.seedance3.tools\u002Fseedance3-tools\u002Fweb\u002Fassets\u002Fblog\u002Fseedance-2-5-prompt-guide\u002Fseedance-2-5-prompt-example.mp4",[14,313,314],{},[140,315,316,317,150],{},"Official Seedance 2.5 timed-prompt example. Source: ",[45,318,149],{"href":100,"target":145,"rel":319},[147,148],[172,321,323],{"id":322},"_2-rainy-neon-detective-scene","2. Rainy Neon Detective Scene",[14,325,326,328],{},[26,327,292],{}," Cinematic storytelling and controlled tension.",[110,330,333],{"className":331,"code":332,"language":115,"meta":116},[113],"A tired private detective in a charcoal coat walks through a rain-soaked neon street at midnight. 0-6s: wide rear tracking shot as he moves through the crowd. 6-15s: the camera draws beside him while he notices a red envelope beneath a flickering sign. 15-24s: he stops, picks it up, and studies the unfamiliar wax seal. 24-30s: close-up as his expression changes from curiosity to concern. Blue and magenta reflections, realistic rain, restrained handheld movement, neo-noir realism. Wet footsteps, traffic, distant thunder, no music. Keep his face, coat, and age unchanged.\n",[118,334,332],{"__ignoreMap":116},[172,336,338],{"id":337},"_3-mountain-travel-reveal","3. Mountain Travel Reveal",[14,340,341,343],{},[26,342,292],{}," Destination videos and tourism campaigns.",[110,345,348],{"className":346,"code":347,"language":115,"meta":116},[113],"Create a 16:9 cinematic travel reveal in the Dolomites at dawn. 0-7s: begin close behind a hiker walking through cool blue mist. 7-16s: the camera cranes upward as she reaches the ridge. 16-25s: reveal golden sunlight crossing the mountain peaks and the valley below. 25-30s: hold a wide hero composition with the hiker at the lower-left third of frame. Natural outdoor realism, gentle wind moving her jacket and grass, subtle lens flare. Include boots on gravel, wind, and distant birds. No dramatic speed ramps and no location text.\n",[118,349,347],{"__ignoreMap":116},[172,351,353],{"id":352},"_4-thirty-second-micro-drama","4. Thirty-Second Micro-Drama",[14,355,356,358],{},[26,357,292],{}," Short emotional stories.",[110,360,363],{"className":361,"code":362,"language":115,"meta":116},[113],"Create a restrained live-action micro-drama in a nearly empty railway station. 0-6s: a young man waits on a bench holding an unopened letter. 6-14s: an elderly woman sits beside him and silently recognizes the handwriting. 14-23s: he passes her the letter; she opens it and smiles through tears. 23-30s: the arriving train fills the background with light while they remain seated together. Use simple shot changes, natural performances, soft winter daylight, and muted cinematic color. Include station ambience and the approaching train. No narration. Preserve both characters' faces, clothing, and positions across every shot.\n",[118,364,362],{"__ignoreMap":116},[172,366,368],{"id":367},"subject-and-character-reference-prompts","Subject and Character Reference Prompts",[172,370,372],{"id":371},"_5-consistent-fashion-character","5. Consistent Fashion Character",[14,374,375,377],{},[26,376,292],{}," Fashion films using several character images.",[110,379,382],{"className":380,"code":381,"language":115,"meta":116},[113],"Use @Image1, @Image2, and @Image3 as front, side, and full-body references for the same model. Preserve her face, short black hair, silver earrings, red coat, and body proportions. Create a fashion-film sequence in a minimalist concrete gallery. She walks forward, pauses beneath a skylight, turns left, and looks back toward the camera. Use a slow full-body tracking shot followed by a controlled medium close-up. Clean editorial lighting, realistic fabric movement, no wardrobe changes, no extra accessories, and no changes to her facial identity.\n",[118,383,381],{"__ignoreMap":116},[172,385,387],{"id":386},"_6-brand-mascot-adventure","6. Brand Mascot Adventure",[14,389,390,392],{},[26,391,292],{}," Keeping a stylized mascot recognizable.",[110,394,397],{"className":395,"code":396,"language":115,"meta":116},[113],"Use @Image1 as the identity reference for the small orange robot mascot. Preserve its round head, two blue eyes, white chest logo, short arms, and matte material. The robot leaves a charging station, crosses a playful miniature city, helps a tiny delivery vehicle over a bump, and finishes by waving beside the original charging station. Bright 3D animation, smooth family-friendly motion, stable scale, cheerful electronic sound effects. Do not redesign the logo, add fingers, change eye color, or introduce another orange robot.\n",[118,398,396],{"__ignoreMap":116},[172,400,402],{"id":401},"_7-consistent-product-hero","7. Consistent Product Hero",[14,404,405,407],{},[26,406,292],{}," Branded product videos.",[110,409,412],{"className":410,"code":411,"language":115,"meta":116},[113],"Use @Image1 and @Image2 for the exact bottle shape, cap, label placement, typography, and amber glass color. Create a premium product reveal on dark volcanic stone. Water moves around the base while a narrow warm light travels across the glass. The camera performs one slow half-orbit and finishes on a centered front view where the label is fully readable. Preserve product geometry and branding throughout. Do not bend the label, duplicate the bottle, change the cap, add text, or cover the logo with water.\n",[118,413,411],{"__ignoreMap":116},[172,415,417],{"id":416},"_8-talking-portrait-from-image-and-audio","8. Talking Portrait from Image and Audio",[14,419,420,422],{},[26,421,292],{}," Dialogue, voice reference, and subtle performance.",[110,424,427],{"className":425,"code":426,"language":115,"meta":116},[113],"Use @Image1 for the speaker's identity, hairstyle, navy shirt, and softly lit office. Use @Audio1 for the exact voice, pacing, and dialogue. Generate natural lip synchronization, restrained eyebrow movement, occasional blinking, and two small head movements that follow the emphasis in the speech. Keep a stable medium close-up with a very slow push-in. Preserve the face and background. No exaggerated gestures, no camera cuts, no subtitles, and no additional music.\n",[118,428,426],{"__ignoreMap":116},[172,430,432],{"id":431},"motion-reference-prompts","Motion Reference Prompts",[172,434,436],{"id":435},"_9-dance-motion-transfer","9. Dance Motion Transfer",[14,438,439,441],{},[26,440,292],{}," Reusing choreography without copying the original performer.",[110,443,446],{"className":444,"code":445,"language":115,"meta":116},[113],"Use @Image1 for the new performer's face, athletic build, and black-and-gold costume. Follow only the full-body choreography, rhythm, and timing from @Video1. Do not copy the original dancer's identity, clothing, or background. Place the performance on a reflective stage surrounded by vertical amber lights. Keep the performer fully visible during fast movements, preserve facial identity and costume, and use a smooth front tracking camera. Synchronize movement to @Audio1 and add realistic foot contact with the floor.\n",[118,447,445],{"__ignoreMap":116},[172,449,451],{"id":450},"_10-sports-movement-reference","10. Sports Movement Reference",[14,453,454,456],{},[26,455,292],{}," Athletic motion and branded sports content.",[110,458,461],{"className":459,"code":460,"language":115,"meta":116},[113],"Use @Image1 for the female runner's appearance and blue uniform. Follow the sprint start, stride rhythm, and arm motion from @Video1 only. Replace the original location with a wet stadium track at night. Begin with a low side view at the starting blocks, accelerate into a parallel tracking shot, and finish with a front three-quarter shot as she crosses the line. Preserve anatomy, uniform, lane direction, and running speed. Add spikes on wet track, controlled breathing, and distant crowd sound.\n",[118,462,460],{"__ignoreMap":116},[172,464,466],{"id":465},"_11-camera-movement-reference","11. Camera-Movement Reference",[14,468,469,471],{},[26,470,292],{}," Recreating shot language while changing content.",[110,473,476],{"className":474,"code":475,"language":115,"meta":116},[113],"Use @Video1 only for its camera path, shot timing, and reveal structure. Do not copy its subjects, setting, color, or style. Replace the scene with a futuristic motorcycle displayed in a dark industrial studio. Match the opening overhead descent, the low circular reveal, and the final fast push toward the headlight. Use @Image1 for the exact motorcycle design and proportions. Maintain clean product geometry, realistic metal reflections, and a premium automotive-commercial look.\n",[118,477,475],{"__ignoreMap":116},[172,479,481],{"id":480},"product-and-style-reference-prompts","Product and Style Reference Prompts",[172,483,485],{"id":484},"_12-luxury-perfume-commercial","12. Luxury Perfume Commercial",[14,487,488,490],{},[26,489,292],{}," High-end ecommerce and paid ads.",[110,492,495],{"className":493,"code":494,"language":115,"meta":116},[113],"Use @Image1 as the exact perfume-bottle reference and @Image2 only for the purple floral color palette. Begin with the bottle emerging through thin mist. Iris petals rotate slowly behind it while a ribbon of translucent purple liquid curves around the base without covering the label. Use macro highlights, a slow vertical camera rise, and a centered final hero shot. Preserve bottle shape, liquid color, cap, label, and logo. Add soft glass tones and a restrained ambient score. No extra products or floating text.\n",[118,496,494],{"__ignoreMap":116},[172,498,500],{"id":499},"_13-sneaker-launch-video","13. Sneaker Launch Video",[14,502,503,505],{},[26,504,292],{}," Product detail, motion, and social advertising.",[110,507,510],{"className":508,"code":509,"language":115,"meta":116},[113],"Use @Image1, @Image2, and @Image3 for the sneaker's side, sole, and rear design. Create a high-energy launch film on a black studio floor. 0-5s: macro shot of the textured sole landing. 5-12s: the shoe rotates as light reveals the side panel. 12-20s: transition into a runner wearing the same shoe for three powerful steps. 20-25s: return to the clean product hero shot. Preserve every panel, lace, logo, color, and proportion. Crisp impact sound, subtle electronic beat, no altered branding.\n",[118,511,509],{"__ignoreMap":116},[172,513,515],{"id":514},"_14-food-commercial-style-transfer","14. Food Commercial Style Transfer",[14,517,518,520],{},[26,519,292],{}," Applying a visual reference without copying its content.",[110,522,525],{"className":523,"code":524,"language":115,"meta":116},[113],"Use @Image1 for the plated ramen and exact ingredient arrangement. Use @Video1 only for its warm tabletop lighting, slow steam movement, and macro commercial pacing. Do not copy the food or tableware from @Video1. Show broth being poured, noodles lifted once, and the bowl returning to a clean hero composition. Preserve ingredient colors and bowl design. Rich natural texture, realistic steam, gentle restaurant ambience, no hands covering the food, and no text.\n",[118,526,524],{"__ignoreMap":116},[172,528,530],{"id":529},"audio-and-dialogue-prompts","Audio and Dialogue Prompts",[172,532,534],{"id":533},"_15-timed-dialogue-scene","15. Timed Dialogue Scene",[14,536,537,539],{},[26,538,292],{}," Coordinating speech and visible action.",[110,541,544],{"className":542,"code":543,"language":115,"meta":116},[113],"Create a stable two-person medium shot in a quiet design studio. 0-6s: the creative director studies a prototype in silence. 6-14s: she looks at her colleague and says, \"The shape is right, but the story is missing.\" 14-22s: the colleague turns the prototype toward the light and replies, \"Then let the light tell it.\" 22-30s: both watch as the surface detail becomes visible. Natural English voices, precise lip synchronization, restrained gestures, consistent eyelines, soft room ambience, no subtitles, and no overlapping dialogue.\n",[118,545,543],{"__ignoreMap":116},[172,547,549],{"id":548},"_16-voice-referenced-brand-message","16. Voice-Referenced Brand Message",[14,551,552,554],{},[26,553,292],{}," Preserving an approved voice and delivery.",[110,556,559],{"className":557,"code":558,"language":115,"meta":116},[113],"Use @Image1 for the presenter and @Audio1 for the approved voice, tone, pacing, and spoken message. Place the presenter in a bright renewable-energy laboratory. Keep a stable waist-up composition while she speaks, then make one slow camera move toward the prototype she indicates. Generate accurate lip movement and subtle hand gestures at the emphasized words. Preserve identity, clothing, and voice. Keep laboratory ambience quiet beneath the dialogue. No captions or background music.\n",[118,560,558],{"__ignoreMap":116},[172,562,564],{"id":563},"_17-music-synchronized-product-montage","17. Music-Synchronized Product Montage",[14,566,567,569],{},[26,568,292],{}," Cutting visual beats to an audio reference.",[110,571,574],{"className":572,"code":573,"language":115,"meta":116},[113],"Use @Audio1 as the timing and rhythm reference. Use @Image1 for the smartwatch design. Create four connected visual beats synchronized to the main accents: wrist raise, screen activation, running-motion close-up, and final product hero shot. Preserve the watch body, crown, band, interface colors, and logo. Use energetic daylight, crisp motion, and clean match cuts. Sound effects may support the beat but must not overpower the reference music. No unrelated products and no interface distortion.\n",[118,575,573],{"__ignoreMap":116},[172,577,579],{"id":578},"storyboard-and-keyframe-prompts","Storyboard and Keyframe Prompts",[172,581,583],{"id":582},"_18-first-to-last-frame-transformation","18. First-to-Last-Frame Transformation",[14,585,586,588],{},[26,587,292],{}," Controlled visual transitions.",[110,590,593],{"className":591,"code":592,"language":115,"meta":116},[113],"Use @Image1 as the exact first frame and @Image2 as the exact final frame. Begin with the abandoned warehouse composition from @Image1. As the camera moves slowly forward, lights activate one row at a time, machinery repairs itself, plants grow along the walls, and the space becomes the clean creative studio shown in @Image2. Finish precisely on the final composition. Make every transformation continuous and physically connected. No cuts, no flicker, and no objects appearing without a visible transition.\n",[118,594,592],{"__ignoreMap":116},[172,596,598],{"id":597},"_19-storyboard-guided-product-tutorial","19. Storyboard-Guided Product Tutorial",[14,600,601,603],{},[26,602,292],{}," Short explainers with controlled shot order.",[110,605,608],{"className":606,"code":607,"language":115,"meta":116},[113],"Use @Image1 as a six-panel storyboard for shot order, framing, and camera direction. Use @Image2 for the exact portable coffee machine. Follow the storyboard sequence: unpack the machine, fill the water chamber, insert the capsule, press the button, show the extraction close-up, and finish with the cup beside the product. Preserve the product design and step order. Clean kitchen daylight, precise hand interaction, clear instructional pacing, realistic machine sounds, no voiceover, and no added labels.\n",[118,609,607],{"__ignoreMap":116},[172,611,613],{"id":612},"_20-multi-keyframe-fantasy-journey","20. Multi-Keyframe Fantasy Journey",[14,615,616,618],{},[26,617,292],{}," Connecting several designed moments.",[110,620,623],{"className":621,"code":622,"language":115,"meta":116},[113],"Use @Image1, @Image2, @Image3, and @Image4 as sequential keyframes. Preserve the same child, yellow raincoat, red umbrella, and watercolor storybook style. Connect the rainy city street, floating-lantern bridge, cloud garden, and sunrise hill through continuous umbrella-driven transitions. Each location should emerge from the previous environment rather than cut abruptly. Keep character scale and clothing stable, use gentle orchestral music and rain that gradually fades, and finish on the exact composition of @Image4.\n",[118,624,622],{"__ignoreMap":116},[172,626,628],{"id":627},"video-editing-prompts","Video Editing Prompts",[172,630,632],{"id":631},"_21-replace-a-background-without-changing-performance","21. Replace a Background Without Changing Performance",[14,634,635,637],{},[26,636,292],{}," Reusing approved footage in a new location.",[110,639,642],{"className":640,"code":641,"language":115,"meta":116},[113],"Edit @Video1. Keep the actor, facial performance, dialogue, body movement, camera position, timing, and foreground table unchanged. Replace only the background with a modern library at dusk, using @Image1 for architecture, shelf layout, and warm practical lighting. Match the original depth of field and light direction so the replacement feels natural. Do not change the actor's identity, clothing, voice, hands, or lip movement.\n",[118,643,641],{"__ignoreMap":116},[172,645,647],{"id":646},"_22-natural-age-progression","22. Natural Age Progression",[14,649,650,652],{},[26,651,292],{}," Controlled local transformation.",[110,654,657],{"className":655,"code":656,"language":115,"meta":116},[113],"Edit @Video1 while preserving its composition, camera motion, background, clothing, and performance timing. Change only the woman's visible age: she begins in her late twenties and gradually ages to approximately sixty by the final frame. Keep the same facial identity, bone structure, eye color, hairstyle, and expression. Add realistic changes to skin texture and hair over time. The transition must be continuous, with no cuts, identity drift, flicker, or sudden face replacement.\n",[118,658,656],{"__ignoreMap":116},[172,660,662],{"id":661},"_23-remove-an-object-and-its-sound","23. Remove an Object and Its Sound",[14,664,665,667],{},[26,666,292],{}," Coordinated visual and audio cleanup.",[110,669,672],{"className":670,"code":671,"language":115,"meta":116},[113],"Edit @Video1. Remove the red delivery scooter that crosses the background from 4-7 seconds and reconstruct the street naturally behind it. Remove only the scooter's engine sound during the same time range. Preserve the pedestrians, storefront reflections, camera movement, dialogue, music, traffic ambience, and total video timing. Do not blur the area or alter the person speaking in the foreground.\n",[118,673,671],{"__ignoreMap":116},[172,675,677],{"id":676},"extension-and-transition-prompts","Extension and Transition Prompts",[172,679,681],{"id":680},"_24-extend-a-nature-clip","24. Extend a Nature Clip",[14,683,684,686],{},[26,685,292],{}," Continuing motion and ambience seamlessly.",[110,688,691],{"className":689,"code":690,"language":115,"meta":116},[113],"Extend @Video1 forward by five seconds. Continue the bee's existing flight path toward the nearest purple flower. Move into a macro close-up as it lands, collects pollen, and lifts off toward the upper-right edge of frame. Preserve the original flower species, sunlight direction, wind, depth of field, camera speed, color grade, wing sound, and garden ambience. The extension must begin without a visible seam or audio change.\n",[118,692,690],{"__ignoreMap":116},[172,694,696],{"id":695},"_25-seamless-transition-between-two-videos","25. Seamless Transition Between Two Videos",[14,698,699,701],{},[26,700,292],{}," Connecting separately generated scenes.",[110,703,706],{"className":704,"code":705,"language":115,"meta":116},[113],"Seamlessly connect @Video1 and @Video2 without changing their existing content. At the end of @Video1, follow the spinning vinyl record into an extreme close-up of its black surface. Transform the circular highlight into the moon seen at the beginning of @Video2, then pull back into the second scene. Match rotation speed, light direction, color, music tempo, and camera momentum across the transition. No fade to black, visible cut, frozen frame, or abrupt audio change.\n",[118,707,705],{"__ignoreMap":116},[14,709,710,711,716],{},"For more visual ideas, browse the ",[45,712,715],{"href":713,"target":145,"rel":714},"https:\u002F\u002Fseedance3.tools\u002Finspiration",[147,148],"Seedance inspiration gallery"," before adapting these templates to your own scenes.",[92,718,72],{"id":719},"common-seedance-25-prompt-mistakes",[172,721,723],{"id":722},"using-too-many-references-without-roles","Using Too Many References Without Roles",[14,725,726],{},"Fifty available reference slots are not a target. Start with the minimum decisive set and tell Seedance what each file controls.",[172,728,730],{"id":729},"describing-a-moment-instead-of-a-sequence","Describing a Moment Instead of a Sequence",[14,732,733],{},"A longer prompt still needs progression. Write what changes from the opening to the ending.",[172,735,737],{"id":736},"stacking-conflicting-camera-moves","Stacking Conflicting Camera Moves",[14,739,740],{},"Do not request a push-in, orbit, crane, handheld shake, and rapid zoom in the same beat. Choose the movement that best reveals the action.",[172,742,744],{"id":743},"forgetting-continuity-locks","Forgetting Continuity Locks",[14,746,747],{},"If identity or product accuracy matters, repeat the few details that must remain stable and list the changes you do not want.",[172,749,751],{"id":750},"mixing-the-task-with-the-reference","Mixing the Task With the Reference",[14,753,754],{},"Say whether a video should control motion, camera, style, or content. Otherwise, the model may preserve elements you intended to replace.",[172,756,758],{"id":757},"treating-platform-limits-as-universal","Treating Platform Limits as Universal",[14,760,761],{},"ModelArk capabilities and settings may not appear identically on every third-party platform. Check the controls available in your current generator.",[92,763,78],{"id":764},"how-to-edit-specific-details-in-a-seedance-25-video",[14,766,767],{},"Seedance 2.5 can edit a specific detail in an existing video while preserving the original composition, camera movement, lighting, timing, and performance. This is different from rewriting a prompt and generating the entire scene again from scratch.",[14,769,770],{},"A strong video-editing prompt identifies four things:",[772,773,774,777,780,783],"ol",{},[42,775,776],{},"The uploaded video to edit.",[42,778,779],{},"The exact detail that should change.",[42,781,782],{},"The elements that must remain unchanged.",[42,784,785],{},"The continuity problems that should be avoided.",[14,787,788],{},"For example, this prompt changes only the subject's visible age:",[110,790,793],{"className":791,"code":792,"language":115,"meta":116},[113],"Preserve the composition, camera position, lighting, and performance rhythm of @Video1.\n\nOnly change the woman's appearance: gradually age her from her twenties to around sixty while preserving her identity. Let the restraint in her expression slowly soften, with tears forming before she smiles.\n\nKeep the video as one continuous shot. Do not change the background, camera movement, timing, clothing, or body movement. No cuts, flickering, facial drift, or sudden age transitions.\n",[118,794,792],{"__ignoreMap":116},[14,796,797],{},"The instruction separates the requested edit from the protected elements. That gives Seedance a clear boundary: change the woman's age and expression, but preserve the existing shot.",[14,799,800],{},[26,801,802],{},"Before editing",[124,804,130,806,136],{"controls":126,"playsInline":126,"preload":127,"width":128,"ariaLabel":805},"Official Seedance 2.5 video before instruction-based editing",[132,807],{"src":808,"type":135},"https:\u002F\u002Fcdn.seedance3.tools\u002Fseedance3-tools\u002Fweb\u002Fassets\u002Fblog\u002Fseedance-2-5-prompt-guide\u002Fseedance-2-5-official-video-editing-before.mp4",[14,810,811],{},[26,812,813],{},"After editing",[124,815,130,817,136],{"controls":126,"playsInline":126,"preload":127,"width":128,"ariaLabel":816},"Official Seedance 2.5 video after changing the subject's age and expression",[132,818],{"src":819,"type":135},"https:\u002F\u002Fcdn.seedance3.tools\u002Fseedance3-tools\u002Fweb\u002Fassets\u002Fblog\u002Fseedance-2-5-prompt-guide\u002Fseedance-2-5-official-video-editing-after.mp4",[14,821,822],{},[140,823,824,825,150],{},"Official instruction-based video editing example. Source: ",[45,826,149],{"href":100,"target":145,"rel":827},[147,148],[92,829,84],{"id":830},"seedance-25-prompt-faq",[172,832,834],{"id":833},"what-is-the-best-seedance-prompt-structure","What is the best Seedance prompt structure?",[14,836,837],{},"Start with the task, assign reference roles, protect the subject, describe the timeline, connect camera movement to action, add visual and audio direction, and finish with constraints. Simple shots can use a short paragraph; longer scenes benefit from timed beats.",[172,839,841],{"id":840},"can-i-copy-these-seedance-prompt-examples","Can I copy these Seedance prompt examples?",[14,843,844],{},"Yes. Replace the subjects, references, dialogue, and brand details. Treat each example as a production template rather than a magic phrase, then refine one variable at a time.",[172,846,848],{"id":847},"how-long-should-a-seedance-25-prompt-be","How long should a Seedance 2.5 prompt be?",[14,850,851],{},"Use the shortest prompt that clearly controls the result. A five-second product shot may need one paragraph. A 30-second reference-heavy scene may need subject settings, time ranges, camera direction, sound, and continuity rules.",[172,853,855],{"id":854},"how-many-references-can-seedance-25-use","How many references can Seedance 2.5 use?",[14,857,858],{},"The official BytePlus guide documents support for up to 50 combined image, video, and audio reference assets in one request. Your platform or generation mode may expose a lower limit, and fewer well-assigned references are often easier to control.",[172,860,862],{"id":861},"can-seedance-25-generate-a-30-second-video","Can Seedance 2.5 generate a 30-second video?",[14,864,865],{},"BytePlus documents single-video generation up to 30 seconds. Available durations can vary by task and platform, particularly when editing, extending, or generating from fixed boundary frames.",[172,867,869],{"id":868},"do-seedance-20-prompts-work-with-seedance-25","Do Seedance 2.0 prompts work with Seedance 2.5?",[14,871,872],{},"The core principles still apply: visible action, clear camera direction, coherent style, audio instructions, and assigned references. For 2.5, add a structured timeline, stronger continuity rules, and explicit task instructions when using editing, keyframes, or extension.",[172,874,876],{"id":875},"where-can-i-use-seedance-25","Where can I use Seedance 2.5?",[14,878,879,880,885,886,150],{},"Seedance 2.5 is coming soon to Seedance3.tools. Visit the ",[45,881,884],{"href":882,"target":145,"rel":883},"https:\u002F\u002Fseedance3.tools\u002Fseedance-2-5",[147,148],"Seedance 2.5 page"," for availability updates. Until then, you can prepare references and test the fundamentals with ",[45,887,890],{"href":888,"target":145,"rel":889},"https:\u002F\u002Fseedance3.tools\u002Fseedance-2-0",[147,148],"Seedance 2.0",[92,892,90],{"id":893},"create-better-seedance-videos",[14,895,896],{},"A useful Seedance 2.5 prompt does not win by being the longest. It wins by making the production logic visible: what each reference controls, what happens over time, how the camera supports the action, what the viewer hears, and what must remain unchanged.",[14,898,899],{},"Start with one of the 25 Seedance prompt examples above. Replace only the creative variables, run a focused generation, and revise the first point of failure instead of rewriting the entire prompt.",[14,901,902,903,908],{},"If you are working with the current model, follow our ",[45,904,907],{"href":905,"target":145,"rel":906},"https:\u002F\u002Fseedance3.tools\u002Fblog\u002Fhow-to-use-seedance-2-0",[147,148],"step-by-step guide to using Seedance 2.0"," to learn the creation workflow before applying the more advanced reference and editing techniques in this guide.",[14,910,911,915],{},[45,912,914],{"href":882,"target":145,"rel":913},[147,148],"Explore Seedance 2.5 on Seedance3.tools"," and be ready to turn these prompt templates into finished videos when the model goes live.",{"title":116,"searchDepth":917,"depth":917,"links":918},2,[919,920,932,933,968,976,977,986],{"id":94,"depth":917,"text":48},{"id":153,"depth":917,"text":54,"children":921},[922,924,925,926,927,928,929,930,931],{"id":174,"depth":923,"text":175},3,{"id":181,"depth":923,"text":182},{"id":188,"depth":923,"text":189},{"id":195,"depth":923,"text":196},{"id":202,"depth":923,"text":203},{"id":209,"depth":923,"text":210},{"id":216,"depth":923,"text":217},{"id":223,"depth":923,"text":224},{"id":230,"depth":923,"text":231},{"id":246,"depth":917,"text":60},{"id":276,"depth":917,"text":66,"children":934},[935,936,937,938,939,940,941,942,943,944,945,946,947,948,949,950,951,952,953,954,955,956,957,958,959,960,961,962,963,964,965,966,967],{"id":282,"depth":923,"text":283},{"id":286,"depth":923,"text":287},{"id":322,"depth":923,"text":323},{"id":337,"depth":923,"text":338},{"id":352,"depth":923,"text":353},{"id":367,"depth":923,"text":368},{"id":371,"depth":923,"text":372},{"id":386,"depth":923,"text":387},{"id":401,"depth":923,"text":402},{"id":416,"depth":923,"text":417},{"id":431,"depth":923,"text":432},{"id":435,"depth":923,"text":436},{"id":450,"depth":923,"text":451},{"id":465,"depth":923,"text":466},{"id":480,"depth":923,"text":481},{"id":484,"depth":923,"text":485},{"id":499,"depth":923,"text":500},{"id":514,"depth":923,"text":515},{"id":529,"depth":923,"text":530},{"id":533,"depth":923,"text":534},{"id":548,"depth":923,"text":549},{"id":563,"depth":923,"text":564},{"id":578,"depth":923,"text":579},{"id":582,"depth":923,"text":583},{"id":597,"depth":923,"text":598},{"id":612,"depth":923,"text":613},{"id":627,"depth":923,"text":628},{"id":631,"depth":923,"text":632},{"id":646,"depth":923,"text":647},{"id":661,"depth":923,"text":662},{"id":676,"depth":923,"text":677},{"id":680,"depth":923,"text":681},{"id":695,"depth":923,"text":696},{"id":719,"depth":917,"text":72,"children":969},[970,971,972,973,974,975],{"id":722,"depth":923,"text":723},{"id":729,"depth":923,"text":730},{"id":736,"depth":923,"text":737},{"id":743,"depth":923,"text":744},{"id":750,"depth":923,"text":751},{"id":757,"depth":923,"text":758},{"id":764,"depth":917,"text":78},{"id":830,"depth":917,"text":84,"children":978},[979,980,981,982,983,984,985],{"id":833,"depth":923,"text":834},{"id":840,"depth":923,"text":841},{"id":847,"depth":923,"text":848},{"id":854,"depth":923,"text":855},{"id":861,"depth":923,"text":862},{"id":868,"depth":923,"text":869},{"id":875,"depth":923,"text":876},{"id":893,"depth":917,"text":90},null,"2026-08-12","Copy 25 Seedance 2.5 prompt examples and learn how to direct longer scenes, references, camera movement, audio, editing, and consistent AI videos.","md","https:\u002F\u002Fcdn.seedance3.tools\u002Fseedance3-tools\u002Fweb\u002Fassets\u002Fblog\u002Fseedance-2-5-prompt-guide\u002Fseedance-2-5-prompt-guide.webp","seedance prompt, seedance prompt examples, seedance 2.5 prompts",{},"\u002Fblog\u002Fseedance-2-5-prompt-guide",{"title":5,"description":989},{"loc":994,"lastmod":988},"blog\u002Fseedance-2-5-prompt-guide","w1CQVyBtM_Z2vShUWmzQILyWuu6rKi2tfH2Yhd8-kIg",1786503017521]