carat

Prompt Gallery

/

Seedance 2.5

Login
Sign Up

Seedance 2.5

Seedance 2.5 by ByteDance turns a single prompt into a 30-second video, with up to 50 references and native audio sync - great for music videos, ads, and cinematic clips.

2026.07.01

·

One Flower Passed Across Nine Countries

Seedance 2.5· Video
Live-action style, fast cutting, cinematic, 4K, 24fps, warm natural light, real human performance, natural lip-sync, no subtitles. The core visual thread of the whole video is the passing of a single fresh flower: the flower travels quickly from one country to the next, linking different regions and people around the globe. In each scene a character receives the flower, shows a sincere smile, and says 'thank you' in the local language. Overall rhythm light and fluid, dynamic camera, emphasizing real street/life atmosphere, cross-cultural warm connection, and the passing of goodwill between people. Transitions: the previous character passes the flower out of frame and the next character catches it in a new scene, or a fast whip-pan, motion blur, or foreground occlusion completes a seamless transition, keeping the flower visually continuous - a 'one shot around the world' feel. Camera style: handheld tracking, slight camera shake, fast push-pull, mixing close and medium shots, real ambient sound, cinematic street-shot texture. Background music warm, light, with a world-travel feel, gently fading out at the end. Scene 1 (@Image1) inside a Chinese flower shop: a girl takes a rose, looks at camera and smiles, saying naturally '谢谢!'; the flower enters from the right and she lifts the bouquet lightly. Scene 2 (@Image2) an English street, cool weather: a man takes a carnation, smiles and nods 'Thank you!'; a whip-pan carries the flower in. Scene 3 (@Image3) a Mexican market, colorful and lively: an older woman takes a bunch of marigolds, presses her hands together, warmly '¡Gracias!'; the camera sweeps past stalls and crowds, freezing on the moment she takes the flower. Scene 4 (@Image4) an Indonesian village, natural sunlight: a child takes a frangipani, smiles happily and bows slightly 'Terima kasih!'; the camera has a running feel, mood innocent and natural. Scene 5 (@Image5) a Thai street, bustling: a vendor takes a jasmine garland, hands pressed together, kindly 'ขอบคุณค่ะ!'; the camera glides forward, the garland swaying gently in the sun. Scene 6 (@Image6) an Arabian courtyard, soft light, elegant: a woman takes a desert rose, hand on chest, smiling; the frame quiet and warm, expression sincere. Scene 7 (@Image7) a Brazilian community, warm and vivid: a boy takes a gerbera, very happy 'Obrigado!'; the camera rhythmic, full of life. Scene 8 (@Image8) a Japanese street: an office worker takes a small flower atop a bento, bows politely 'ありがとう!'; the shot short and crisp, keeping urban rhythm. Scene 9 (@Image9) a Korean street, modern city feel: a young woman takes an azalea, hands naturally together, smiles '감사합니다!'; the camera lingers on her smile a moment, then the frame gently fades out.
@johnAGI168

Brand Concept Film Through Windows and Eyes

Seedance 2.5· Video
A cinematic brand concept short. @Image1 is the first frame; the picture shakes slightly, the camera gradually pushes in, arriving at tree shadows rushing backward outside a window, the shadows receding faster and faster, then suddenly cuts to @Image2 where the speed abruptly slows; the camera moves gently forward along a stream, birdsong and flowers. The camera moves down, arriving underwater, with the sound of bubbles (@Image3 @Image4); a group of orange jellyfish drift gracefully across the frame. The camera slowly pulls back; a school of small fish flit across and pass from the water into the window. A girl looks left and right, watching the little fish. The camera slowly pulls back, the image goes out of focus, then refocuses and becomes clear, switching in time with the music: a Chinese-garden lattice window (@Image5) with light turning, a church stained-glass window, an airplane cabin window, a domed skylight, a black maple window, a louvered blind, a European dormer window, a peephole in a door, a camera viewfinder frame, a bird's eye, a human-eye close-up. The frame rests on the human-eye close-up, then the eye closes, the screen goes black, then suddenly the eye opens and the word 'seedance' appears in the center of the eye with an accent.
@johnAGI168

One-Take Walk Through Six Style-Shifting Rooms

Seedance 2.5· Video
One continuous shot (oner). The camera smoothly follows a person in a black coat (reference @Image1) moving left to right through six connected rooms of different color palettes and moods. Every room shares the same structure: white walls, herringbone light-wood floor, French double floor-to-ceiling windows, white sheer curtains (reference @Image2), but the view outside each window and the indoor mood are completely different. The protagonist walks at a steady pace throughout, passing through every open door in the walls. 0-5s, Room 1, theme American-comic fight: the protagonist enters and fights a character (@Image3); the character is defeated. 5-10s, Room 2, theme warmth, felt/wool-craft style; outside the window is a sunflower field (@Image4), indoor light warm-orange and soft, a painter is painting sunflowers (@Image5). After entering, the protagonist also turns into felt style. 10-15s, Room 3, theme sorrow: the whole frame becomes a black-and-white manga stop-motion style; rain outside, cold-grey low indoor light, one person sits alone on the floor in the center of an empty room, head down hugging their knees, a phone beside them glowing with an unanswered call screen. The protagonist turns off the room light on entering, then immediately turns it on - the room bursts into color and a roomful of fresh flowers grows in an instant. 15-20s, Room 4, theme joy: the entire scene is a room submerged in the sea (reference @Image6); the protagonist swims in, beautiful coral reefs and schools of fish beside them. 20-25s, Room 5, theme surprise: outside the window a sky full of fireworks at night (reference @Image7), colorful flickering light reflecting indoors, the protagonist swept into the celebratory mood. 25-30s, finally the protagonist arrives in a blank white room, stands in the center and snaps their fingers; a finger-snap sound effect plays, the frame goes fully black, and the word 'seedance' appears in the center (reference @Image8). Overall cinematic texture, high-fashion advertising style; the light is entirely determined by the scene outside each window, creating strong emotional contrast. No text in the frame.
@johnAGI168

Cinematic Youth Racing 2D Animation Short

Seedance 2.5· Video
A 30-second cinematic youth racing drama short in 2D animation style. The protagonist is a teenage rider competing in a high-level motorcycle race. Overall style: hot-blooded, youthful, emotionally intense, highly cinematic, with a complete narrative arc and a clear emotional curve. Only two camera moves are used throughout: high-speed tracking and slow-motion orbit. Minimal dialogue, appearing naturally like fragments of memory; tone sincere, gentle, restrained - no slogans, no overwrought sentiment. No sense of disaster, no negative expression, no exaggerated sci-fi. Focus on love, support, comeback, and growth within a youthful race. 0-5s: A dusk racetrack, an intense high-speed opening. The camera hugs the ground, tracking the boy's motorcycle at high speed, tires grazing the track edge, engine roaring, wind sharp, atmosphere tense and blazing. The boy is fully focused; the setting sun draws sharp highlights on the metal shell of the bike. 5-9s: After a key corner the boy is suddenly overtaken by a rival. The high-speed tracking pushes on, conveying the pressure of a dropping rank and broken rhythm. A close-up inside the helmet shows a brief daze, tightened breathing, a slight waver. The boy murmurs: 'Can I still catch up...' 9-14s: The boy falls behind, breathing heavier, emotions sinking to a low. The race has not stopped; the bike still races forward. Amid the ride, warm memory fragments flash back: as a child learning to ride, someone steadied him from behind; his father adjusting his helmet, careful and quiet; before the finish line, a gentle smile watching him; on a dusk slope, silhouettes walking side by side. These memories are shown in golden backlight, soft slow motion, fragmented feeling. 14-18s: The music turns from suppressed toward soaring. Restrained, gentle voices come from memory: 'Don't be afraid - I'm always here.' 'Stay steady. And look forward.' The boy's eyes refocus, breathing slowly steadies, emotion shifts from wavering to resolute. 18-23s: The boy regains his conviction, accelerates with everything, cuts a precise line to fight back. High-speed tracking shows the power and control of leaning into and out of corners, closing in on the bike ahead. Low and firm, the boy says: 'I won't stop here.' 23-27s: An upward stretch of track appears ahead; the boy sprints at full speed into the sunset. Only breathing, engine, and rising music remain. The bike launches into the air on its momentum, entering stunning slow motion. From deep in memory, a last gentle, smiling voice: 'Go on.' 27-30s: The camera orbits the airborne bike in slow-motion close-up, pushing the hot-blooded, tender, free, upward-leaping emotion to a climax. Flowers bloom behind in a special effect, then 'seedance' appears (reference @Image1).
@johnAGI168

Anime Idol Concert Stage Performance

Seedance 2.5· Video
Cinematic anime idol concert CG, large-scale live stage performance, 2.5D anime rendering with realistic stage lighting and volumetric beams, 15 seconds, continuous high-energy singing and dancing, no pauses, no freeze frames, no story beats. Atmosphere of Love
@Mayz1169

Continuous Static Shot Counting to Ten

Seedance 2.5· Video
One continuous static shot, no cuts. A person looks directly at the camera and counts from one to ten, holding up the corresponding number of fingers as they say each number.
@patrickassale

TikTok Shop Live Stream Screen Recording

Seedance 2.5· Video
screen recording of a TikTok Shop live stream featuring a woman selling products.
@patrickassale

Chibi Boxer Doll Animation from Storyboard

Seedance 2.5· Video
Use @Image 5 as storyboard to drive the whole animation. Use @Image 1 only for the boxer doll's final identity: chibi felt-and-paper doll, big pale wool afro with dark undercut, thick drawn eyebrows, round blush cheeks, tattooed
@byarlooo

Ancient Chinese Character Biography Short Film

Seedance 2.5· Video
【任务】 生成完整 30 秒中国古风人物小传短片。 严格按照时间轴、多模态素材映射、人物隔离和参考音频执行,不自由修改镜头与剧情。 【参考绑定】 - Depth 视频:只控制约 0–5s、10–18s、25–30s 三段核心舞蹈 - 多视角头像 + 全身图:只锁定主舞女
@MrLarus

Handmade Clay Mug 3D Animated Short

Seedance 2.5· Video
[SCENE] A 30-second stylised 3D animated short. A handmade clay mug at the back of a kitchen cupboard waits to be chosen, while the perfect porcelain cups at the front are used every day. Warm, funny, quietly moving. 2.35:1.
@Dheepanratnam

Ancient-Style Costume OOTD Short Video

Seedance 2.5· Video
7秒,9:16,一镜到底,甜美古风OOTD短视频。7秒音频完整播放,配合人物舞蹈。 严格以@人物参考图人物设定图中的成年女性为唯一人物:以右侧脸部特写作为人物长相参考,以左侧正面图 and 中间背面图作为服装、发型及背部结构参考。完整保持她的小巧鹅蛋脸、清秀五官、红色眉心花钿、深棕黑色高盘古典发髻、浅金色花叶发簪、珍珠流苏和两侧垂落珠链;完整保持浅香槟白与淡桃粉古风舞衣、刺绣抹胸内搭、浅杏色透明长袖纱衣、浅金色腰封、金属链饰、多层垂坠长裙。全程禁止换脸、换装、改变发型或增减饰品。 仅参考@视频人物的动作轨迹、姿态、节奏、景别变化和构图,不继承原视频人物的脸、发型、服装、颜色或材质。纯白无影棚,高机位轻微俯拍,超广角轻微透视,镜头基本固定,人物始终保持全身或大半身清晰入镜。 0–1秒,人物从画面右侧提着少量裙摆轻快走入,长裙和薄纱袖自然摆动; 1–2秒,人物靠近镜头,上身轻微前倾,抬眼注视镜头,神情清甜克制; 2–3秒,双手轻扶腰间,形成自然收腰姿态,俏皮微微歪头; 3–4秒,优雅侧身转过半圈,展示服装背面、盘发和腰封结构,再回眸看向镜头; 4–5秒,转回正面浅浅微笑,双臂向身体两侧自然舒展,轻纱袖缓慢垂落; 5–6秒,双手轻放身后,先低头整理姿态,再微微歪头看向镜头; 6–7秒,再次轻扶腰间,人物先扭头看向与镜头呈90°角的侧前方,视线先聚焦在镜头外某个远方,然后再转向镜头,以甜美、自信、端庄的古风姿态定格结束。 动作自然连贯、轻盈克制,不做大幅度舞蹈动作。裙摆、薄纱衣袖、珠链流苏和耳侧发丝随动作产生轻微真实惯性,避免夸张飞舞。明亮柔光,清透高亮,柔和浅香槟色调,古装影视广告与韩系时尚短片结合的高级质感,节奏轻快,无对白,搭配轻盈空灵的国风流行音乐。人物身份、五官、发型、服装纹样和饰品位置全程稳定,面部清晰稳定,真实皮肤纹理,4K质感,细节清晰。 不要使用左侧正面图中的空白遮脸作为人物面部,人物长相必须来自右侧脸部特写;不要继承视频中原人物的脸、发型或服装;不要新增人物,不要换装,不要跳切,不要镜头大幅移动或抖动,不要快速旋转,不要剧烈甩袖,不要裙摆遮脸,不要肢体变形、重复手臂、多余手指、手部粘连、穿模、服装突变、饰品消失、发型变化或背景变化。
@acupofai

Rap Performance with On-Screen Text

Seedance 2.5· Video
A main character performs a rap song while “yes&” appears on the screen. 6-second video.
@yesand_ai

What Is Seedance 2.5?

Seedance 2.5 is ByteDance's next-generation video model that turns a single prompt into a 30-second video in one pass. Where most AI video models cap out around 15 seconds, Seedance 2.5 generates one continuous scene without stitching short clips together.

It accepts up to 50 multimodal references - images, video, audio, and style inputs (up from 12 in the previous generation) - so you can control the subject, tone, and composition far more precisely. Sound is generated in the same space as the visuals, so onscreen action and audio stay in sync.

Seedance 2.5 became available through its public API in August 2026. It isn't available in Carat yet, but you can create video today with Seedance 2.0, which uses the same video grammar.

Where You'll Use It

  • 30-second long-form clips from a single prompt
  • Image-based video conversion (i2v) that brings a still photo to life
  • Music video and K-pop stage clips built from multiple references
  • Ads and UGC videos with automatically synced sound
  • Cinematic and action clips with controlled camera movement

Why This Tag Is Useful

Seedance 2.5's strength is generating longer video in one continuous take - with sound included. A full 30 seconds of setup, development, and payoff can be designed in a single prompt, cutting out the old workflow of making and stitching separate clips. Locking a subject, product, or mood with up to 50 references keeps the same face and tone throughout, and because motion and audio are generated together, effects like footsteps and ambient sound land in sync. For now, Seedance 2.0 is the closest model you can run in Carat with the same grammar (@Image1 references, Shot and timecode multishots, tracking and dolly moves), so practicing there is the best way to get ready for 2.5.

Tips for Building Your Prompt

  1. For a full 30 seconds, split the scene into setup, development, and payoff using Shot 1: / Shot 2: or 0.0-2.5s timecodes.
  2. Add reference images for any subject, product, or style you want to keep, and point to them with @Image1, @Image2, and so on - up to 50 inputs.
  3. If you need sound effects or ambience, describe which sound matches which action so the audio syncs with the visuals.
  4. State the aspect ratio (16:9 · 9:16 · 1:1) and length up front, and control motion with camera moves like tracking, dolly, and push-in.

Until Seedance 2.5 is supported, get familiar with the same grammar in Seedance 2.0. In Carat you can generate video with Seedance 2.0, then keep going with follow-up work like changing the ratio, adding captions, and upscaling.

Frequently Asked Questions

Related Prompts

Movie/Drama

487 prompts

Handheld Video

38 prompts

Character

Character

363 prompts

Dialogue Video

Dialogue Video

24 prompts

Carat Prompt Gallery

Create trending AI prompts instantly

Prompts

  • Logo
  • PPT Design
  • Thumbnail
  • Poster
  • Web Novel Cover
  • ID Photo
  • Caricature
  • Product Detail Page
  • Banner

Models

  • GPT Image 2
  • Seedance 2.0
  • NanoBanana Pro
  • NanoBanana 2

Carat

  • Get started
  • About
  • Blog
  • Careers
  • Contact
  • Partnership
  • Terms
  • Privacy

© 2026 Paradot.Inc All rights reserved.