YouMind
Sign in

2D Animation

Explore free 2D Animation video prompts from YouMind's AI prompt library, grouped under styles. Every prompt is curated from real creative workflows and ready to copy, adapt, and reuse.

397 promptsDaily highlights
Claude Opus 5.5
View full prompt & copy

<direction> Build a 30-second looping motion design showreel that introduces CLAUDE, a motion designer. Concept, in one line: one pencil line gets promoted three times (graphite on paper, wire on a blueprint, light on black) and never leaves the screen. Every shot is made out of the previous one: shapes, type and colour transform into the next scene. No cuts, no fades. Typeface: Geist only. Weight 900 for big words, 600 to 800 for small labels (24 to 28 px, tracked 0.14 to 0.3 em). Palette: paper cream #EDE6D5, blueprint blue #1B3ED2, hologram cyan #45F0DD, plus black #07080A, graphite #16161A and white. Each stage's accent is the next stage's colour: blue pencil on paper, cyan on the blueprint, white on the hologram. Camera: locked and front-on for the whole film, like a drafting table. Objects turn; the camera never moves. Empty border: nothing at rest within 150 px of the sides or 90 px of the top and bottom. Things travelling through may cross it. Three inks. The same geometry is drawn in whichever ink is on screen: - PAPER: a cream paper background (use my paper image if I give one, otherwise flat cream with fine drawn grain) with a 60 px blue grid at about 13% and every fifth line at 26%. Lines are pencil: resampled and pushed off true by slow sine waves whose phase changes 8 times a second, stroked twice (full width at 90%, half width at 40% with more wobble), with paper tooth knocked out of the strokes, then multiplied onto the paper. Fills are diagonal hatching drawn on one line at a time. Big type is hatching cut to the letter shapes, with a doubled outline. - BLUEPRINT: flat blue with the same grid in white. Clean white lines 3 to 4 px, hidden edges dashed and thinner, fills are fine clean hatching. - HOLOGRAM: black with the same grid in cyan. Cyan lines with a white core 30% of the line width. Fills are horizontal scan lines on an 8 px pitch whose thickness follows how lit the surface is. No blur halo anywhere: it reads as light through line density only. Banned: particle bursts, glows, lens flares, camera shake, RGB split, crossfades, generic gradients, bouncy easing, holds longer than half a second, anything that looks like a template. </direction> <structure> 120 BPM, 60 beats, 15 bars of 2 seconds. Something happens on every beat. Moves start 0.05 s before their beat so they land on it. Easing: a critically damped spring, 1 - (1 + w t) e^(-w t) with w between 12 and 40, so nothing ever passes its target; exponential ease-in and ease-out for wipes, presses and exits. Scene changes: paper to blueprint by a shape that floods the frame with a hard edge; blueprint to hologram and hologram to paper by one horizontal scan line. The last frame is the first frame, so it loops. Two corner labels stay on all film: top left "FIG. 07 — GEAR" (bar number and name), bottom right "BEAT 27 / 60". Bar names: DRAFT, TYPE, CURVE, MORPH, TILE, SOLID, GEAR, PANEL, DATA, LIGHT, WAVE, GLOBE, DRUM, FORM, END. </structure> <mascot> The creative director: a small box character. Body 76 x 54 units with rounded corners, two short arm stubs (24 x 16) at mid height, four straight legs (11 x 33), two tall rounded-rectangle eyes (8 x 14). Draw it about 250 px wide. Its right arm is raised 58 degrees holding a pencil. On paper: graphite outline with sparse blue hatching and graphite eyes. In the hologram: cyan outline with scan-line fill and white eyes. It is never a fourth colour. It runs with alternate leg pairs lifting, six steps a second, leaning 7 degrees into the run. </mascot> <storyboard> Bar 1, HOOK, 0-2 s, paper. Frame 0: CLAUDE in Geist 900, fitted to 1500 px wide, centred, pencil-hatched. INTRODUCING small above it, A MOTION DESIGNER small below. A pencil rail 1620 px wide sits 96 px above the letters. The mascot stands on the rail at x = 1560 with CREATIVE DIRECTOR lettered to its left. 0.0: the small lines un-type; the mascot runs off the right edge, accelerating. 0.36 to 0.5: the rail drops, accelerating, to 26 px below the baseline and presses the letters flat into itself. 1.0: the line rises to the middle, shortens to 1320 px and splits into two rails 140 px apart; 41 blue ruler ticks sweep along the top rail. 1.5: the rails open to 300 px; a blue height dimension draws at the right with a live number. Bar 2, KINETIC TYPE, 2-4 s, paper. Words sit between the rails and are clipped to that slot; the rails move to fit each word. 2.0: INTRODUCING rises through the baseline letter by letter, then lifts out through the top. 2.5: A MOTION whips in from the left (1400 px in 0.2 s). 3.0: the rails clap shut on it and reopen, stretching DESIGNER out of the line; its letters then fly out sideways. 3.5: CLAUDE in blue pencil, letters dropping from above one every 0.05 s. At 4.0 the rails close and crush it into one line 1476 px long. Bar 3, CURVES, 4-6 s, paper. 4.1: the line gets pen-tool handles (430 px long at 58 degrees, small squares at the ends, circles on the handle tips, the label EASE) and bends into an S. 4.5: the handles swing to the mirror angle; eight blue copies fan out 34 px apart, joined at both ends like a leaf. 5.0: handles retract; the ribbon becomes travelling sine waves (amplitude 92 px, 2.75 cycles, each line a little out of phase). 5.5: it gathers back into one flat line. Bar 4, SHAPE MORPH, 6-8 s, paper. 6.1: the line curls into a circle of radius 235 (curvature grows, length stays). 6.5: triangle, turning 120 degrees as it changes. 7.0: square. 7.5: hexagon, point up, radius 270. Blue hatching builds inside, then cross-hatching. Bar 5, TILING, 8-10 s, paper. 8.0: the hexagon shrinks to radius 96 and six more slide out around it. 8.5: twelve more, a honeycomb of 19. 9.0: dense hatching sweeps across until every cell is solid blue. 9.6 to 10.0: the centre cell's shape grows past the frame with a hard edge; inside it is the blueprint, with the honeycomb in white line. Bar 6, 2D TO 3D, 10-12 s, blueprint. 10.0: three lines draw inside every hexagon, plus three dashed hidden ones: each is a cube seen corner-on. 10.5: the eighteen outer cubes slide into the centre one, which grows to radius 270. 11.0: it turns a quarter and perspective switches on. 11.5: it faces front. Bar 7, MECHANISM, 12-14 s, blueprint. 12.0: the cube flattens and its front square grows teeth: a 16-tooth gear, pitch radius 208. From 12.25 the train steps 45 degrees on every eighth note. 12.5: a 10-tooth gear grows out to the upper right, meshing. 13.0: a 12-tooth gear to the upper left. Each has a dashed pitch circle, a hub and a label (16T, 10T, 12T). 13.5: the small gears retract; the gear becomes a square again and the cube gets its depth back. Bar 8, UNFOLD, 14-16 s, blueprint. 14.0: the cube unfolds on its hinges into a cross of six squares. 14.5: the squares slide and stretch into six panels: a header strip, a tall sidebar, a wide chart, two tiles, a timeline strip. 15.0: they fill in. Header: "CLAUDE — MOTION REEL", a BARS / RING switch, a LIGHT toggle. Chart: 30 bars, one per second of this film, heights 42 30 36 44 50 46 52 62 50 56 62 72 60 66 58 76 56 62 68 74 70 78 86 94 100 93 97 90 80 50. Timeline: the 15 bar names, a keyframe diamond on each. 15.5: counters roll up: 120 BPM, 60 BEATS, 1800 FRAMES, 0 CUTS. Bar 9, DATA, 16-18 s, blueprint. 16.0: a graph line draws across the bar tops; a cyan playhead lands on the timeline at the real time. The current second's bar is solid cyan, earlier ones hatched, later ones outlined. A cursor arrives. 16.5: it clicks RING: the chart's baseline bends round into a circle of radius 110 with the ba

Generate video
Seedance 2.5
View full prompt & copy

Rules: - Theater anime quality with a budget of 500 million yen scale, Neo-Retro flat Japanese 2D anime. Vivid cyan-blue outlines, gentle pastel colors - One shot (no cuts). Camera is fixed and does not move until the end. Do not use fisheye lens - High number of drawing frames. Fast tempo, comical - No BGM, sound effects present - No subtitles, no text on screen {{img 1}}: Subject - Maka {{img 2}}: Subject - Roni [Consistency] Do not change Maka-chan's face, pink skin, short pink bob, large cream-colored round ear-like head accessories on both sides, large blue eyes, white round earrings, red choker, yellow puff-sleeve top, red belt, and white skirt throughout the video. Do not change Roni-kun's face, pink skin, curly pink hair, large cream-colored round ear-like head accessories on both sides, blue eyes, mint green blazer, purple and yellow checkered shirt, and white pants. Only Maka-chan and Roni-kun appear. Take out one item at a time. Do not duplicate items or Roni-kun. [FRAME-BREAK] MAIN VISUAL GOAL: An unmistakable OUT-OF-FRAME, POP-OUT effect. Maka repeatedly reaches beyond the central picture window. Her hands, forearms and every object she pulls out visibly move OVER the black letterbox bars, toward and past the viewer. COMPOSITING ORDER, BACK TO FRONT: 1. Background and scene. 2. Two broad black letterbox bars at the top and bottom. 3. The foreground characters: Maka (her hands, arms, pink hair, cream ear-like head accessories and every object she holds) and, at the end, Roni as she pulls him out. The background is confined to the central picture window. The foreground character is NOT confined to that window. Her silhouette continues uninterrupted across its boundaries and onto the black bars. Her pink skin, pink hair, cream accessories, yellow sleeves and each object remain fully visible when overlapping black. STATIC BARS, MOVING CHARACTER: Both bars are completely static screen graphics, not objects in the scene. Keep their equal thickness, full width and exact screen position unchanged throughout. Their inner boundaries remain perfectly straight and horizontal. The character simply renders in front of them. Her movement temporarily hides portions of their black surface. The underlying bars never move or deform. Visible edge segments on either side of an overlap remain aligned on the same horizontal line. When she withdraws an arm, reveal the original unchanged black bar beneath it. Keep black visible beside the overlapping arms and objects. No bending, sliding, shaking, stretching, opening or impact reactions in the bars. The camera stays stationary. One continuous shot. She stays close and large in the composition, leaning toward the viewer. Never retreat deep into the background. The frame-break effect must come from the character visibly covering black areas outside the central picture window. Do not keep her entirely behind the bars. Do not simulate the effect by moving the bars or enlarging the window. [Meaning] Beyond the black margin below (outside the bottom of the screen) is this side = the viewer's room. Maka leans out from the edge of the screen, puts her hand in towards this side, and looks for something. [Generation Goal] Maka rummages around this side of the screen, throws various items back one after another, and finally pulls out Roni-kun who was hiding on this side. (6 stages, one continuous shot without breaks. The stage remains the same) Location: A sparse room with only plain light cream walls and light mint floor. No furniture, windows, shelves, or long horizontal objects. Light is soft and uniform, shadows are thin. Items thrown back by Maka fall on this floor and accumulate gradually. Maka is right in front of the camera, head to chest filling the height of the window largely. She does not retreat into the background. [S1] From the start, Maka's face is large within the window. She leans forward strongly, peering at us with big blue eyes. With one hand, she places fingertips on the black upper margin as if gripping the edge (peering while gripping the edge). Tilting her head, "something..."

Generate video
Claude Opus 5.5
View full prompt & copy

Make a 32-second video, 1920x1080, 30fps, about one idea: AI is built on human brilliance. Every model learned from people who knew things. LOOK: a printed street-poster collage, in motion. No photos and no stock images: draw everything yourself with SVG and CSS. - Warm off-white paper (#F2EFE8) with a fine dotted halftone texture over the whole frame, and a very slow paper drift so the frame is never frozen. - Black ink line art in an old engraving style (hatching lines, not flat fills): a human eye, a pair of hands, a flower, and a neuron with branching dendrites. Lines draw themselves on (stroke-dashoffset), then a halftone dot fill fades in under them, like a printed collage layer. - Flat lavender paper blocks (#A79BF0 and a deeper #6E5FD8) that slide in behind the art with slightly rough edges, offset like torn paper. - One real formula set in a serif math face that writes itself on, symbol by symbol: the sum from n=1 to infinity of 1/n^2 = pi^2/6. - Headlines in one heavy geometric sans (bold, tight tracking), black text on white bars, big and poster-like. Every line fully inside the frame with generous margins. STORY, three beats, use these lines exactly: Beat 1, THE SPARK (0-10s): the eye draws itself in at the center, lavender blocks slide behind it. Text: "Before the machine, a mind." then "Someone noticed. Someone asked why." Beat 2, THE WORK (10-22s): three collage cards cut in on the beat, each with its art and one line: "A proof at midnight." (the formula writes itself), "A hand that heals." (the hands), "An ear for the right word." (the flower opens). Then: "Every answer it gives, people taught it first." Beat 3, THE CLAIM (22-32s): the neuron grows its branches outward across the frame, connecting the eye, hands, formula and flower into one poster layout. Text: "Human brilliance is the input." then "Everything else is compute." Hold the final poster with a slow push-in until the last frame. SOUND: if the app has a music library, add one calm, minimal piano or ambient track under the whole video, fading out at the end; add a soft paper-slide sound when blocks move in, if sound effects are available. Motion must start on frame 0 and keep going to the very last frame. No logos, no brand names.

Generate video
Claude Opus 5.5
View full prompt & copy

<inputs> Ask me for: a short label for the first toggle, one main word, a percentage for the graph, text for the final card, one accent color and a track with a clear beat. </inputs> <direction> A 2D motion design film in 16:9. Use a warm off-white canvas, near-black backgrounds, bold type and one vivid orange accent. The orange dot links every scene: it begins in the UI, becomes the dot above a letter, draws a spring graph, expands into a shape and reappears on the end card. A cursor makes the opening click feel real. Keep the movement crisp and controlled, with subtle spring overshoots and no empty pauses. Banned: stock footage, gradients, glows, particle bursts, arbitrary transitions and movement that doesn't grow from something already on screen. </direction> <structure> Around 15 seconds, with a visual change on every major beat. Open: a cursor clicks the orange toggle beside "Reduce Motion". The letters bend and shift into place until only "Motion" remains, with the orange dot above the "i". Spring: the dot leaps away from the word and lands on a graph. Its trajectory draws a spring response, passes the target and settles next to a "+22%" callout. Shape: the dot races along a line, expands into an orange circle and fills the frame. A black rounded square rotates inside it, then extends into a pill. Everything contracts into one orange dot on black. Type: words including "SPRINGS", "EASING", "KINETIC TYPE", "3D" and "SOUND" circle the dot. The ring of words spreads into a field of tiny dots. Reveal: the halftone field expands and turns black, uncovering the final card: "Opus 5.5", "Motion designer." and "0 keyframes". Return: the orange dot travels back through the halftone, graph and toggle. Make the last frame match the first for a seamless loop. </structure> <build> 1. Use one HTML file, 1920x1080 at 60fps. Calculate every visual from time inside seek(t): no CSS transitions, timers or state carried between frames. 2. Use closed-form springs for the toggle, moving dot and graph. Draw the graph with the same spring function that controls the dot. 3. Move the letters individually as "Reduce Motion" becomes "Motion". Keep the orange dot above the "i" until it jumps away. 4. Treat the orange circle, rotating square and pill as states of one shape. Carry its position and momentum through each change. 5. Arrange the words on a circular path, then send their letters outward to transition into the halftone grid. 6. Draw the halftone as a dot grid whose sizes react to the orange dot. Expand it beyond the frame before showing the end card. 7. Align the click, dot landing, graph peak, shape changes and final title with measured beats. Render one frame per beat to check timing before the full export. </build>

Generate video
Claude Opus 5.5
View full prompt & copy

<inputs> Ask me for: a target value for the slider, three short notification messages, a reply to type, one accent color and a royalty-free track around 120 BPM. If I skip anything, use 100, "Slider - Reached 100", "Gravity - Something is falling", "Messages - What now?", "retry" and electric blue. </inputs> <direction> A 10-second 2D UI motion film in 16:9. Place everything on a near-black canvas with a faint dot grid. Use charcoal controls, sharp white text, one blue accent and a white ball that moves through the entire film. A real cursor makes the opening click feel physical; the ball guides each transition that follows. Keep the UI clean and the impacts precise, with only small spring overshoots. Move the camera through one continuous layout so every interaction fills the frame. Banned: hard cuts, stock footage, glows, particle bursts, disconnected transitions, excessive bouncing and blurry text during camera moves. </direction> <structure> Around 10 seconds, with something happening on every major beat. Toggle: the cursor clicks a blue switch. Its white ball springs free, and the camera follows it toward a horizontal slider. Slider: the ball lands on the knob and moves it along the track. The blue fill stays attached, while a value bubble above the knob climbs to 100. Notifications: hitting 100 sends the ball onward and brings in three dark cards: "Slider", "Gravity" and "Messages". Each arrives on its own beat. The camera follows the ball down to "What now?". Reply: the small "Reply" pill stretches into a text field. A keyboard rises below it as part of the same scene. Keyboard: the ball strikes the keys while "retry" appears in the field. It lands on Return, the blue send button activates and the reply contracts into a small blue message. Return: the camera pulls back to show the toggle, slider, cards and keyboard together in one world. The ball travels back to the first switch. Match its position, the cursor and the switch state to the opening frame so the film loops cleanly. </structure> <build> 1. Make one HTML file at 1920x1080 and 60fps. Calculate every visual from time inside seek(t): no CSS transitions, timers or state carried between frames. 2. Arrange all UI components in one world. Use a single camera transform with position and log-space zoom, making each close-up a move through that world. 3. Give the white ball one continuous route between the toggle, slider, cards and keys. Use closed-form springs for landings and a brief squash on impact. 4. Calculate the slider fill and number from the ball's horizontal position. Keep the tooltip centered over the knob until it reaches 100. 5. Give every key press a fixed time. Show the reply as the portion typed by that moment, so any frame can be rendered deterministically. </build>

Generate video
By @Vox
Claude Opus 5.5
View full prompt & copy

Make a 3 to 4 minute animated explainer, 16:9, about natural gas extraction for general public, in a flat illustration style, with narration and captions that match it: 1. Open on something the audience sees every day, then go back to the beginning. 2. Follow the clearest line: steps for a process, otherwise a timeline, cause and effect, or problem to solution. 3. One chapter per step, 8 to 12 chapters, titles in the top-left corner. Keep one visual style and palette throughout. 4. Pick one number that runs through the whole video (temperature, pressure, speed, cost or time), keep it in a corner, and update it at every step. If no number fits, show "step X of Y". 5. Give every key number an everyday comparison, and leave out any number you're not sure of. 6. End with one line that ties the whole journey together. Before you start, send me the chapter list, the number you picked and the key facts you'll use, so I can confirm.

Generate video
Seedance 2.5
View full prompt & copy

A lone pixel-art bounty hunter rides a skeletal motorcycle across a toxic desert while giant sand worms burst from beneath the highway, she swerves through collapsing asphalt, jumps over a burning tanker, fires an energy harpoon into the largest creature and lets it drag her across the dunes before using the chain to slingshot herself onto its head; cinematic high-detail pixel art, rich 32-bit aesthetic, multilayer parallax desert, fluid sprite animation, dramatic sprite scaling, heavy screen shake, heat distortion made from pixel dithering, orange sunset, toxic green storms, cyan weapon trails and flying debris, brutal retro-futuristic atmosphere, ending with the hunter planting a charge between the worm’s armored plates and jumping back onto the motorcycle as the creature erupts into a gigantic pixel explosion behind her.

Generate video
Claude Opus 5.5
View full prompt & copy

Create an interactive HTML presentation explaining the concept of 'Network Effects' where connections grow exponentially as user count increases (e.g., 5x users = 27x connections). The interface should include a slider allowing the viewer to manually adjust the number of users and see the resulting change in connection density in real-time. Use only HTML and JavaScript without external image assets. Include voice narration synchronized with the visual changes.

Generate video
Claude Opus 5.5
View full prompt & copy

<inputs> Give me: * canvas color * shape color * accent color * shape radius * fonts * the states the shape should morph through Use hex codes, seconds, damping ratios and pixel sizes instead of adjectives. </inputs> <direction> Make a 14 second, 1440x1440, 60fps seamless UI motion loop. One shape, zero cuts. The shape morphs from state to state and is never replaced. Each state fills 50% of the frame. The camera starts on the beat and settles slower than the shape. Use closed-form springs. Shape damping 0.72, settle time 0.35s. Camera damping 1.0, settle time 0.6s. Content swaps use 12px blur and fade over 0.15s. The cursor tip stays locked to whatever it drags. </direction> <structure> Use a 0.5s beat grid. 0.0 - pixel mascot idles 1.0 - mascot squashes into play button 1.5 - loader 2.0 - check draws 2.5 - timeline scrub 4.0 - keyframe stretches past the end and snaps back 5.0 - bezier editor 7.0 - counter rolls F 000 to F 840 8.5 - render bar fills 10.0 - command palette types "export mp4" 11.0 - toast says "840 frames rendered" 12.0 - toast shrinks back into the mascot The last frame must be the first frame. </structure> <build> Write one HTML file with Canvas 2D. window.renderAt(t) must be a pure function of time. No Math.random. No clock. No carried state. ?t= must freeze the exact moment. Render every frame from renderAt(frame / 60). Use 4 subframes for motion blur. Encode to H.264 at 60fps. </build> <gotchas> Fix the cursor jump at the loop wrap with: t - 14*floor(t/14) Do not cross-fade orange into ink. Use a ripple from the center at 1.5s and from the click point at 12.0s. Roll word swaps vertically. Start draw-on after blur-in. Aim the cursor at where the target will be when it lands. Keep cursor speed under 65 px/frame. </gotchas> <start> First show me the storyboard. Then write the code. After the first render, check 5 frames around every beat and fix the issues before the final render. </start>

Generate video
Claude Opus 5.5
View full prompt & copy

<输入项> 请向我询问:我的产品、用户在其中输入的确切请求、用于结果卡片的6到8张照片、一首约130 BPM且具有清晰高潮的免版税歌曲(来自Mixkit,可免费商用),以及用于结尾卡片的品牌标志。如果我跳过任何内容,请使用默认设置:下方的点餐故事、来自Pexels的食物照片(火鸡面、Birria塔可、辣味炸鸡、水牛城鸡翅、 Smash汉堡、辣味拉面),以及Mixkit的“Cat Walk”。 </输入项> <指导> 一段15秒的产品发布影片,分辨率1920x1080,帧率60fps。仅限2D,单次连续拍摄。无淡入淡出、模糊或剪辑:每个场景都从上一个场景中生长出来。文字从遮罩线升起,按钮扩展成页面,颜色从一个物体涌入并在收缩成另一个物体前扩散。给相机每个场景一个缓动移动,并且不要连续进行放大和缩小。 外观:暖白色画布 (#f5f5f2),黑色UI (#0e0e10),石灰绿强调色 (#cdf24f),橙色 (#ff6a2a),以及蓝色发送按钮 (#2f6bff)。在输入框周围添加类似Apple Intelligence风格的虹彩光晕:一个模糊的圆锥渐变,旋转穿过粉色、橙色、石灰绿、青色和紫色。在深色卡片上使用真实的食物摄影,UI使用Geist字体,结尾卡片上的产品名称使用衬线字体。 禁止事项:交叉淡化、模糊进入、3D、粒子效果、方向反转的相机移动、超过1秒的停顿,以及任何感觉像模板的内容。 </指导> <结构> 130 BPM,32拍(14.8秒)。在高潮前12拍开始播放音乐。 第0-6拍,提问:缓慢推近发光的输入框,大约3倍缩放。以自然的按键节奏在三行固定文本中输入请求: “我真的饿了,我不知道 我想要什么,但想要一些 辛辣、酥脆且管饱的东西” 第6拍,发送:按下发送按钮。一道石灰绿的涟漪从中扩散,并将消息转化为聊天气泡。拉远一次以展示周围的手机(428x900屏幕,11px黑色边框,9:41状态栏)。 第6-12拍,思考:发送按钮长成一个橙色的炒锅,一个带有细手柄的半圆盘。它在闪烁的文字“正在为你寻找……”下方,在第8、9、10和11拍抛出6种彩色食材。 第12拍,高潮:一个石灰绿色的圆圈从炒锅扩大,在约0.35秒内填满画面。深色食物卡片(330x440,圆角半径28,顶部为照片,名称,一行描述,价格和评分)从中心爆发出现。滑动条带穿过三张卡片并停留在“火鸡面”上,它随着一个“最佳匹配”标签一起抬起。 第16-21拍,聊天:石灰绿色收缩回用户的消息气泡中,携带其文字,同时选中的卡片飞入其槽位。“找到了。来自Wok & Fire的火鸡面,烟熏、辛辣且非常管饱。”升起,随后是“要我帮你下单吗?”然后显示两个药丸状按钮:“立即下单”为黑色,“更多选项”为轮廓样式。光标点击立即下单 → 加载动画 → “✓ 已下单”。 第21-32拍,结尾:黑色按钮扩展成页面,保持药丸形状直到触及边缘,同时相机推进其中。其标签增长至标题大小,然后滑走。逐字引入“你当时想吃面条。”和“现在它已在路上。”(92px,黑底白字)。一个快递员点沿着路线跟随到一个图钉。图钉变成品牌标志,文字标识从其后方擦除消失,上方写着“由...制作”,下方写着产品名称。 </结构> <构建> 1. 一个HTML文件,1920x1080。在seek(t)内部根据时间计算所有样式:不使用CSS过渡、计时器或帧间状态。 2. 在容器上使用单个相机变换,键值为[time, zoom, x, y],具有缓动段,且缩放在对数空间中插值。 3. 对所有的弹出和静止使用闭式阶跃响应弹簧。 4. 使每一次交接成为一个共享元素:石灰绿洪水携带气泡文字的副本,按钮将其标签带入页面。 5. 声音:为每个事件使用下载的Mixkit音效 - 打字声、发送爆破声、每次抛掷声、高潮冲击声、卡片呼啸声、点击声、成功音调、页面填充呼啸声和标志闪光声。根据其测量峰值放置每一个,并将响度标准化为-14 LUFS。 6. 使用Playwright以60fps渲染,每帧使用8个子帧,并通过ffmpeg tmix混合。逐帧审查每个快速时刻并扫描完整视频的每一帧。 </构建>

Generate video
Claude Opus 5.5
View full prompt & copy

The card should scale in from 30% to 100% scale and transition from invisible to visible. Every detail should slide in orderly, one after the other, once the card settles at its final scale. The graph line should animate in like a line path. The glow at the back should become visible as the card details slide in. The two rectangles should move across the screen at slightly different timings to create a light effect.

Generate video
Claude Opus 5.5
View full prompt & copy

<inputs> Ask me for: a voiceover or song with a spoken line (the audio file) and its transcript, 10-12 small objects that sum up my life, two silhouettes to light like a heat camera (a head in profile and a raised hand), and the 3 words I want to land hardest. If I skip any, use the defaults: vinyl, book, camera, cat, coin, game controller, cap, heart, money, plant, clapperboard, skateboard; free CC0 silhouettes from Openverse; action, intention, curiosity. </inputs> <direction> A 20 second 4:3 film, 1440x1080 at 60fps, cut to the voice. Off-white paper, black, one red, one yellow, Geist for every word. Film grain on everything, a soft vignette, and a camera that never rests: a slow handheld float under everything plus a slow push inside every scene. No scene cuts to the next: each handoff starts on the exact last frame of the scene before. The only hard cuts are the beat flips on the last word. Every object is pixel art you draw yourself: a 16x16 grid per object, every cell a hand-placed colour, rendered as a 3D block (a darker extrusion down-right, a thin bevel on every cell) with a soft contact shadow. The silhouettes are lit like a heat camera: the heat comes from how thick the body is at each point, so the nose, the hair and the fingertips run cooler, plus one hotspot and a slow moving noise, mapped through a black > red > orange > cream ramp. The edge stays razor sharp, with a soft glow outside it. Motion rules: nothing pops in (eased glides of about 0.5s with a soft start), fast moves get real motion blur instead of ghost copies, and captions type with the voice behind a block cursor. Banned: stock footage, AI video, 3D camera moves, particles, glitch. </direction> <structure> Hang every scene on the transcript's word timings (measure them with faster-whisper word timestamps, one short chunk at a time): 1. The first word fills a yellow card. The card wipes off to the left and the question types underneath on two centred lines that glide to stay centred as they grow, the key word underlined in red as it's spoken. 2. The objects streak past in lanes above and below the words, a cyan four-point flare sweeps in, then the page burns away: black closes in from the edges onto the head, behind a hot red edge. 3. The head heats up on black while the answer types low on the left. A thin white orbit draws itself around the head, its far half passing behind it. On the last word, paper blooms out from inside the head with the objects in it, and the camera falls through: the head rushes past the lens until the paper is all we see. 4. The objects sit on a tilted ring seen in perspective (the front ones bigger and lower, drawn last), slowly turning, each one bobbing; the coin flips and lands with sparks. 5. On each of the 3 words, that object flies out of the ring to the left and grows, the page floods black from it, and the word types beside it (camera: a red shutter star; book: a red ribbon sweep; record: two white orbits and red notes). Then the flood drains back into the object as it flies home, and the camera dives closer into the ring. Put each hero object in the slot that is front-left when its word lands. 6. The ring pulls in a touch, bursts outward, every object turns to ink, and ink blots grow and swallow the page. 7. The hand rises into frame, hot, with the line split either side of it. The heart drops into its palm, then lifts to the centre at the size the next scene holds it. 8. The last word, one letter per beat, the background flipping paper / black / red, the heart holding its place across the cut, the centre object morphing pixel by pixel into the next one. 9. Thin brush strokes sweep through, the letters drift apart and float, then settle back into the word with the heart above it. The camera keeps pushing through the hold. </structure> <build> 1. One HTML canvas. Every frame is a pure function of time inside seek(t). 2. Bake each pixel sprite once to an offscreen canvas (the extrusion layers, then the face), then draw it scaled and rotated. 3. The heat: precompute each silhouette's base heat once (a distance transform of the mask, held at the edge value outside it so nothing dark bleeds in when it is upscaled). Every frame, add the hotspot and two octaves of value noise at quarter resolution, map it through a 256-step lookup, upscale it, and cut it with the full-resolution mask. 4. The fall through the head: draw the ring scene into an offscreen layer, mask it with the head scaled up about a point inside the skull, and grow that scale on a cubed curve until the head covers the frame. The ring inside zooms less than the head does, so it reads as depth. 5. The pixel morph: pair every cell of object A with a cell of object B and glide each one's position and colour, with a small random delay per cell. 6. Render with Playwright at 60fps with 8 motion-blur subframes, then lay the original audio back on with ffmpeg. </build> <gotchas> A mask PNG with no alpha channel fills the whole box with heat, so put the silhouette in the alpha. A limb cut by the photo's border shows as a straight edge, so extend the mask past the frame. A handoff only reads as smooth when the last frame of one scene is the first frame of the next: draw the outgoing scene under the transition, and give scenes that show the same ring one shared zoom function. A flood that's still retreating when the next scene starts reads as a hard cut, so let it finish first. A fast ease-out flood reads as a jump; use a soft start. </gotchas> <start> Ask me for the inputs, measure the word timings, then show me 8 stills before you render. </start>

Generate video
Loading more prompts

FAQ

Where do YouMind prompts come from?
YouMind curates high-signal AI prompt examples from creator communities and public creative workflows. The library focuses on prompts that are useful to inspect, reuse, and adapt.
How are video prompt categories organized?
Video prompts are grouped into use cases, styles, and subjects. Use cases cover the production goal, styles cover motion or rendering language, and subjects describe the main scene or entity.
How are prompts assigned to categories?
YouMind uses AI to analyze each video prompt, its title, metadata, and generated result when available, then assigns the closest categories. Editors can refine categories as the taxonomy evolves.
What is YouMind?
YouMind is an AI workspace for turning research, ideas, and creative references into usable outputs. The prompt library is one public part of that workflow, built to help creators learn from strong examples faster.