One idea.
Many shapes.

When speech mentions revenue, profit, or a monetary result.

When speech summarizes a method in three steps.

When comparing a previous result with a new result using two numbers.

At the end of a video, when asking viewers to tap a button.

When answering a frequently asked question.

When explaining four steps in sequence.

When listing mistakes, tips, or reasons.

When comparing two options side by side.

When demonstrating a sequence of actions inside an app.

When describing a customer conversation that led to a sale.

When presenting monthly results.

When describing sales arriving one after another.

When explaining an estimated mortgage payment.

When presenting a property with a price and dimensions.

When listing symptoms and possible contributing factors. Replace illustrative content with verified clinical information.

When identifying body regions and possible causes of pain. Illustrative content does not replace medical advice.

When explaining how to divide a plate into food groups.

When suggesting a food substitution.

When explaining a running legal deadline. Verify the applicable jurisdiction and facts.

When quoting a legal provision. Replace illustrative legal text with the applicable verified source.

When celebrating or discussing a follower count.

When replying to a viewer's comment.

When showing a legal article on a document. Verify the jurisdiction and cited rule.

When explaining a legal time limit on a timeline.

When adding up the components of a customer refund.

When explaining the importance of the first three seconds.

When explaining successive audience tests for a video.

When listing viewer engagement signals.

When introducing heart rate and cardiovascular health.

When explaining pressure in a clear or narrowed artery.

When illustrating a risk range or an undiagnosed condition. Verify clinical figures before use.

When revealing the sugar content of a packaged drink.

When translating grams of sugar into cubes.

When comparing sugar with a daily limit and suggesting whole fruit. Verify the intended guidance.

When introducing a property development. The building rises floor by floor while its floor count animates; set floors to zero to hide the count.

When explaining a location and walking distance. The count can land on a spoken word or an explicit panel time; up to two additional locations light up when named.

When showing a property interior: an outlined floor plan becomes a 3D view with area and four room labels.

When introducing a property price above the floor plan. Zero hides the numeric price.

For a standard scene change: a warm light leak peaks at the cut.

For a sharp, emphatic cut: a short white flash peaks at the cut.

For a change of topic or time: film-burn shapes sweep across the screen.

For technology, errors, or sudden reversals: separated color channels, displaced strips, line tearing, and static peak at the cut.

For a section ending or a joke: the image closes into a line and opens again.

For a new item or topic: a strip in the accent color sweeps across the screen.

To move toward an important detail: speed rays and blur emphasize the cut.

For a change of setting or day: seven strips close and open in sequence.

To close a story or focus on a detail: a circle closes at the center and reopens.

For a dramatic pause or passage of time: the image briefly fades to black.

To point at a button, number, product, or link.

To draw attention to a detail with a hand-drawn curved arrow.

To point toward nearby text, a button, or an on-screen item.

To reject an option or mark something as incorrect.

To confirm, approve, or mark an item as completed.

To circle a number, face, button, or other detail.

To underline an important word, number, or phrase.

To outline a region, screenshot detail, button, or object.

To highlight a written phrase, value, or screenshot line.

To isolate a face, product, or screenshot detail by darkening the surroundings.

To conceal an area such as a face, name, number, or screenshot detail.

When asking viewers to enable or disable a feature or mode.

When asking viewers to tap or click a specific point.

When numbering steps or items over the footage.

To surround an important object, face, or number with emphatic rays.

When asking a question or showing uncertainty near the speaker's head.

When warning about an error, risk, or important precaution.

To create a darker cinematic mood or draw attention to the speaker's face.

To blur the background while keeping the face sharp, like portrait mode.

For a dream, memory, luxury mood, or beautiful moment: bright areas gain a soft halo.

For an old-lens, dreamlike, or dizzy look: red and blue fringes appear around edges.

To emphasize a striking number, reversal, surprise, or impact with camera shake.

When discussing sales, profit, revenue, or incoming money.

To celebrate a milestone, sale, achievement, or birthday with confetti.

For storytelling, revelations, or epic moments with a film-trailer look.

For a serious scene, memory, drama, or secret with a monochrome look.

For an old CRT, surveillance camera, archival recording, or memory.

For a weak signal, interference, failing memory, or suspense.

For a nostalgic recording, 1990s home video, or archival VHS look.

For dreams, dizziness, amazement, or confusion: prism-like image copies spread apart.

To give a phrase physical weight: stone letters hang briefly, then fall and break the floor.

For a screen-hit joke: glass cracks at the selected point, the image shakes, and the cracks remain.

For a request for attention or light: the room darkens and a stage spotlight shines on the subject while side beams sweep.

For a concert or nightclub joke: the room darkens, a white strobe pulses to the beat, colored beams cross, and fog rises.

For an explosion joke: realistic fire, dark smoke, and sparks erupt at x/y, with a white flash and a short shake.

For a heat or fire joke: realistic flames rise from the lower edge and a warm flicker covers the footage.

For a disappearing act, mishap, or concealed change: gray smoke rises, covers the screen, and dissipates.

For a broken-screen joke: glass cracks at x/y and the shards fly and fall, with a short flash and shake.

For a strong headline: text blocks alternate weights and the starred keyword glows in the corner-light color.

To challenge apparently good advice: a green card is struck by a light streak, breaks apart, turns red, and reveals a verdict.

When a metric rises or falls: a white curve draws progressively while its marker and counter move to the final value.

To show something that never quite finishes: a progress bar stops near completion, locks, and reveals an emphatic phrase.

When success follows repeated attempts: failed attempts receive red crosses and strikes; the successful attempt lights up green.

When listing completed work or benefits: dim items light up one by one with green checks on a glass card.

When introducing an object with numeric facts: a rotating 3D object appears with a 2×2 fact card. Change the video src to a bundled object file; the renderer reads that HTML resource directly.

When explaining a build or staged process: a 3D house assembles while step labels light up. Playback runs once at 1.2×, completing in 2.4 seconds; allow at least 3.5 seconds for the badge to follow completion.

When discussing reach or locations in Brazil: an angled map lights states individually, by region, or in a chosen set, while a counter animates above.

When introducing the universe, a central idea, or orbital relationships: four planets orbit a 3D Sun, passing behind and in front of it.

When showcasing video work: six portrait cards fan out in 3D and one moves forward. Replace placeholder gradients with muted video elements; set data-start, data-duration, and the correct src. Optional views count below.

When listing requirements, steps, or tasks: a title types in and each item gains a check in sequence.

When showing a request to AI: the request types in, sends, waits, and receives a reply with a finished video.

When inviting a search or introducing a product: the query types in and a leading result appears.

When discussing filming or appearing on camera: the footage becomes a camera viewfinder and a phrase appears word by word.

When showing a computer or AI command: the command types in, Enter submits it, and stages receive checks one by one.

When explaining cost and elapsed time: the cost card counts up, slides aside, and gives way to a time card.

When listing tasks that are no longer necessary: each item appears and is crossed out.

When comparing an old recurring cost with a new price.

When describing a low or high cost, effort, or risk: a gauge indicates the level.

When stating how long a task takes: a counter reaches the chosen time and unit.

When showing editing in progress: an editor panel displays footage, waveform, captions, a moving playhead, and successive cuts.

When explaining removed pauses: red waveform gaps disappear one by one, the waveform joins, and the duration decreases.

When correcting brushing technique: hard brushing marks the tooth red, the wrong title is crossed out, and a soft brush gently cleans a protected tooth.

When explaining muscle recovery: training contracts the muscle; sleep, protein, and rest then send energy into it and it grows.

When explaining retained and lost rights: retained items gain checks while excluded items are crossed out and move away. Verify the applicable jurisdiction and facts.

When comparing whole fruit with juice: oranges are squeezed into one glass, sugar remains in the juice, and fiber moves out, with labels.

When explaining a small-room design: window light reflects from a mirror, then a floor-to-ceiling curtain demonstrates height.

For an opening year or number hook: large digits flip over a dark chart panel and a pill badge emphasizes the message.

When presenting several professions or niches: stacked cards fan out with an occupation icon on each.

At the closing call to action: a comment box appears, a keyword types in, and a finger taps Send.

For standard short-video captions: white words remain visible while the current spoken word turns yellow and bounces.

For a hook or emphatic reversal: one or two large words pop on their spoken timing; the last word is highlighted.

For energetic technology, AI, or editing captions with a chromatic glitch look.

For flowing captions: the current word rises from behind a line with a yellow glow, then turns white.

For a warning, promise, or strong claim: dim words turn red as spoken and remain lit until the next caption group.

For sports, action, challenges, or results: italic captions emphasize the longest word in each group.

For tutorials or busy backgrounds: a subtle dark translucent pill contains the phrase, with a brighter current word.

For conversational tips, routines, or behind-the-scenes footage with a delicate glowing outline.

For commands, code, or AI tutorials: letters type into a dark strip with a moving block cursor that blinks during pauses.

For an opening headline: short black text appears on colored strips, one strip per line near the top.

For a classic short-video hook: a small white lead-in sits above a large accent-colored second line.

For an emphatic number, price, or promise: text jumps up, sharpens, and leaves a large translucent echo. The entrance lasts 0.5 seconds, then holds until the cut.

For an energetic short callout: text springs from zero, overshoots, and settles. It then holds until the cut.

For a callout held for several seconds: letters appear individually and wobble like hand-drawn lettering through four stepped poses.

For a sustained call to action: letters shift and their outlines twist around a wavy baseline in four poses per cycle, faster and stronger than Doodle.

For a technology headline or launch notice: text appears with a chromatic jolt, holds, then briefly splits red and cyan channels across displaced bands.

For an emphatic exit: text flashes on under a moving spotlight; the spotlight closes toward the center and disappears in 0.3 seconds.

For an expiring offer or deadline: text types in, holds, then ripples, flattens, blurs horizontally, and evaporates in a 0.5-second exit.

For numbered lists: a top pill holds the current number and item name. Use one copy per item, with adjacent copies meeting at their boundaries. Editor captions remain visible below; do not hide them.

When introducing an image, screenshot, or clip with impact: the card drops from above, holds, then exits left in the last 0.3 seconds. Set image to a file in the image folder. For footage, replace the image element with a video retaining its ID and class, and add src, data-start="0", data-duration, data-track-index="2", muted, and playsinline. Keep portrait cards beside or below the face; the default 16:9 card sits above it.

When introducing an image, screenshot, or clip with impact: the card drops from above, holds, then exits right in the last 0.3 seconds. Set image to a file in the image folder. For footage, replace the image element with a video retaining its ID and class, and add src, data-start="0", data-duration, data-track-index="2", muted, and playsinline. Keep portrait cards beside or below the face; the default 16:9 card sits above it.

When introducing an image or result: the card slides in from the right, holds, then drops downward in the last 0.3 seconds. Set image to a file in the image folder. For footage, replace the image element with a video retaining its ID and class, and add src, data-start="0", data-duration, data-track-index="2", muted, and playsinline. Keep portrait cards beside or below the face; the default 16:9 card sits above it.

When introducing an image or result: the card slides in from the left, holds, then drops downward in the last 0.3 seconds. Set image to a file in the image folder. For footage, replace the image element with a video retaining its ID and class, and add src, data-start="0", data-duration, data-track-index="2", muted, and playsinline. Keep portrait cards beside or below the face; the default 16:9 card sits above it.

When an image needs a forceful reveal: it enters at 2× scale, snaps down with radial blur, holds still, then rapidly grows at the end. Entry and exit each take about 0.2 seconds. Set image to a file in the image folder. For footage, replace the image element with a video retaining its ID and class, and add src, data-start="0", data-duration, data-track-index="2", muted, and playsinline. Keep portrait cards beside or below the face; the default 16:9 card sits above it.

When an image unfolds on screen: it starts at 0.5× inside a card surrounded by mirrored copies, snaps to full scale with radial blur, holds, then shrinks again. Entry takes about 0.2 seconds; exit about 0.15 seconds. Set image to a file in the image folder. For footage, replace the image element with a video retaining its ID and class, and add src, data-start="0", data-duration, data-track-index="2", muted, and playsinline. Keep portrait cards beside or below the face; the default 16:9 card sits above it.

When an image enters from the side and recedes: a mirrored edge accompanies a blurred half-screen slide, followed by gradual shrinking and a faster final shrink to about 0.8×. The side parameter selects direction. Set image to a file in the image folder. For footage, replace the image element with a video retaining its ID and class, and add src, data-start="0", data-duration, data-track-index="2", muted, and playsinline. Keep portrait cards beside or below the face; the default 16:9 card sits above it.

When opening a dashboard, link, or result: an image rises at 44% scale, a cursor clicks it, and it expands with an elastic shake and brief blur. It holds until the cut. The 1.3-second entrance compresses below 2.4 seconds to occupy 55% of the scene; sound timing must compress with it. Set image to a file in the image folder. For footage, replace the image element with a video retaining its ID and class, and add src, data-start="0", data-duration, data-track-index="2", muted, and playsinline. Keep portrait cards beside or below the face; the default 16:9 card sits above it.

For before/after proof: original footage covers the screen with a recording label, then a bright handled curtain sweeps across, briefly reverses, and reveals the edited footage beneath. The sweep starts after 1 second (earlier below 2.5 seconds) and takes about 1.8 seconds. Match the original clip to the spoken content; remove its CSS filter to retain original color. Set image to a file in the image folder. For footage, replace the image element with a video retaining its ID and class, and add src, data-start="0", data-duration, data-track-index="2", muted, and playsinline. Keep portrait cards beside or below the face; the default 16:9 card sits above it.