Skip to main content

How to Make a Product Launch Video with Opus 5.5 (6 Prompts That Work)

Zhangcan Ding
Zhangcan Ding · Growth Marketing at iArt.ai · LinkedIn
Updated October 16, 2026

When Opus 5.5 came out, product launch videos were the first thing people made with it. Founders typed one prompt and posted a 20-second motion-graphics launch film that would have cost a studio a week.

We read the 282 prompts creators published alongside their videos. Most of the launch videos came from a handful of ways of writing the prompt, repeated with small changes. Below are the six that worked best for us, a film we made with each one in Opus 5.5 for a real product, and a version of the prompt you can run on your own product in one click. iArt runs Opus 5.5 in its Animation mode, so you can try every prompt here without a terminal or an API key.

Results vary from run to run, so treat each film as what the prompt can do rather than a guarantee. The Linear, Spotify, Notion, Wise and Duolingo films are unofficial concept videos we made to show each prompt. They aren't affiliated with or endorsed by those companies.

How to write an Opus 5.5 launch video prompt

The opening line decides what kind of launch film you get, and one line is already enough for a good one. For a more professional film, follow it with a short brief, in this order:

[One or two sentences: what the film is, how long, which format.]

IDEA: the one subject or device that carries the whole film.

0–2s: what is on screen, what moves and how, the exact words.
2–5s: …
(one line per beat, to the end)

LOOK: background and palette (hex codes), type, textures, composition.
MOTION: where it is fast, where it holds, how one shot hands over to the next.
SOUND: the music (tempo, instruments, mood) and a sound for each action. Say "No narration." if there is none.

Every prompt below is written this way: the first lines are the creator's style, and the rest is the brief we wrote from our film, with brackets for your product. In blind side-by-side comparisons on iArt's Animation mode (an AI model watched both films each way round), the full brief beat the opening sentence alone for five of the six films we tried, and tied on the sixth. It takes longer to make and uses more credits, because there is more to build. The quickest way to write one is to describe a film you like, second by second.

What makes an Opus 5.5 launch video work

We ran the same product through different opening lines, and the wording decided the kind of film:

  • Permission beats instructions. "Make a video for Linear" gave us a polite product tour. "Like it's your showreel for a résumé. Go all out." gave us light cutting the logo, a camera flying through it and a 3D board, from the same model.
  • One emotional word writes a story. "Make it emotional" was enough for the model to invent a plot with a twist.
  • A shot list gets you exactly what you wrote. The most detailed prompt in the list scripts every 0.6 seconds, and the film follows it closely. Use it when you already know the film you want.
  • Short wins. The launch videos that spread were 15–20 seconds, with something new on screen every second. Meta's own guidance for mobile video is to deliver the message in 15 seconds or less (Meta for Business).

6 product launch video prompts for Opus 5.5

1. The "go all out" showreel

The prompt asks for a launch video "like it's your showreel" and ends with "go all out." That framing tells the model to show off: logo reveals, camera moves, morphing shapes. The product is a stage for motion rather than a feature list.

The prompt (adapted from @hqmank's original):

Make a dynamic 20-second motion graphics launch video for [your product] ([one line on what it does]), like it's your showreel for a résumé. Go all out.

IDEA: The film is built from the product's own interface language (item tags, status icons, a keyboard shortcut, cards on a board, metric cards), cut into a fast, precise dark-mode sequence on the beat. 16:9.

0–2s: a pinpoint of light stretches into a thin line, rotates 45 degrees and splits into diagonal stripes that slice a white sphere; the camera dives through into whiteout.
2–4.5s: dark space full of small item ID tags drifting in 3D parallax. Center type snaps in: "[one line on what it does]", the last words shifting from accent color to white. A diagonal blade wipes the frame.
4.5–7.5s: one glowing status icon morphs on each beat through five labelled states: dashed circle, hollow ring, half-filled amber pie, three-quarter green pie, accent disc with a check that fires a ring of burst ticks. Zoom into the check.
7.5–10.5s: keycaps ⌘ and K appear; a command menu snaps open ("Type a command or search..."), a few typed letters filter it, an option is chosen, and it closes into a small confirmation toast.
10.5–13.5s: [your product]'s main screen as a board in perspective: cards lift with shadows and reorder across columns. Bottom left, three short verbs build one by one, the last in the accent color.
13.5–16.5s: four metric cards float in isometric perspective: a line chart climbing to 85%, progress bars filling, a counter dropping from 4 to 0, a badge flipping from green to accent. They tilt back and push into camera.
16.5–20s: soft radial light beams. The [your product] logo mark appears and slides left to reveal the [your product] wordmark; "[one line on what it does]" fades up beneath, then a pill with the web address. A sheen sweeps the wordmark; fade to black.

LOOK: near-black slate #0B0C10 to #1A1B23 with vignette and dim glows. White, one accent in your brand colour, amber #F2C94C, green #4EBA86, gray #8A8F98. Neo-grotesque sans, monospace IDs. Dark rounded panels, 1px translucent borders, soft shadows. Hero moments centered; boards and metrics on angled grids.

MOTION: fast and disciplined, locked to about 122 BPM. Sharp acceleration, steep deceleration, no overshoot. Transitions are depth pushes, diagonal slices and modal pop-ins, never plain fades.

SOUND: minimal driving electronic track, warm bass pulses, crisp percussion. A laser rise on the opening slice, clicks on each status change and a chime on the check, key clacks and a confirm pop in the menu, quick slides for the cards, ticking counters and a rising arpeggio on the metrics, a deep sub-bass boom on the logo. No narration.
Show full prompt ▾Show less ▴

Try this prompt on iart.ai →

When to use it. Best for developer tools, brand launches and anything you want to look expensive. Not for explaining how a product works; it prioritizes wow over clarity.

2. The narrated explainer

A fixed five-scene structure: the customer's problem, what you do, how it works in three steps, one proof point, your name. We asked for a voiceover that walks through it. Narration fits this format because the video exists to explain — it's an explainer video in miniature; on the emotional story further down, it would kill the moment.

The prompt (adapted from @alex_prompter's original):

Adopt the role of an expert motion designer. Build a 30-second animated explainer for my business. 5 scenes: the customer's problem, what I do, how it works in 3 steps, one proof point, and my name at the end. Bold text, smooth transitions, my brand colours. Add a friendly voiceover that walks people through it, not an ad read.
My business: [what you sell, who it's for, your colours]

IDEA: one customer's task, followed from a stuck moment to a finished result, through three numbered steps inside [your product]'s real interface.

0–3s: white stage, faint dot grid. Lines snap in: "[setup line]", bold "[the problem]", then one large word in your brand colour: "[the pressure, e.g. Tomorrow.]"
3–5.5s: the problem as a simple app card (e.g. an empty document, a clock ticking toward midnight); a tilted brand-colour pill badge slams on: "[the pain in two words]".
5.5–7.5s: a brand-colour circle grows from the badge to fill the frame; your logo and wordmark land.
7.5–10.5s: white words pop in one by one: "[what you do, in one sentence]". Thumbnails of your output float around the edges in soft 3D.
10.5–12s: the colour shrinks into a card on the dot grid; "Here's how it works."
12–17s: badge "(1) [step 1]". [your product]'s input box; a cursor types [a realistic example request], a file chip drops in, the cursor clicks the main button.
17–21.5s: "(2) [step 2]". Source cards get checkmarks one by one; dashed trails flow into a card that fills line by line.
21.5–25s: "(3) [step 3]". Crash zoom into one piece of the result, then pull back to [your product]'s full main screen.
25–27.5s: proof: one chat request and the result transforms on the spot (e.g. bullets become a growing bar chart). Caption: "[one proof point]".
27.5–31s: blur wipe to white. Big type: "[payoff line in two halves]". The finished output cascades into a grid.
31–36s: quick punch-in on the first piece, then a brand-colour circle wipes to the end card: logo, wordmark, [your website], "[tagline]", thumbnails floating around it.

LOOK: white with a light-grey dot grid; one saturated brand colour as colour field and accent, near-black text. Bold geometric sans, extra-bold sentence-case headlines. Cards, browser frames, pill tags, soft shadows.

MOTION: snappy, short holds for reading. Elastic overshoot on badges, cards and words; quick zooms. No hard cuts: circle wipes, scale-downs, pushes, zoom blurs.

SOUND: upbeat mid-tempo tech-pop. Clock ticks and a slam on the badge; a whoosh on every wipe; key clicks, a file-drop pop and a button click; a ping per checkmark; a rising chime on the transform; a final chord. Narrate it: a friendly voice walks the viewer through each scene, like an explainer, not an ad.
Show full prompt ▾Show less ▴

Try this prompt on iart.ai →

When to use it. Best for explaining a product to someone who has never seen it. Not for audiences who already know you; they'll skip the setup. In iArt's Animation mode, choose a voice in the message box before you send it and the film is narrated.

3. The shot-by-shot brief

The most detailed opening of the six: art direction, palette, a timeline that says what happens every second, and rules for motion ("no particles, no lens flares, no camera shake"). The film follows it closely, so this is the prompt for when you already know what you want.

The prompt (adapted from @brainextends's original):

Create an 18-second product film for [your product] inside a dark rounded stage.
Art direction: [background], [text colors], one accent [hex], one clean sans-serif.
Timeline:
0–1s: the logo arrives with minimal overshoot.
1–2.5s: the headline "[headline]" reveals line by line.
2.5–5s: the product UI rises into view and settles.
5–9s: [the one feature to show], step by step.
9–12s: [the hero moment], large and cropped by the stage.
12–15s: [supporting features] as floating cards.
15–18s: logo lockup with "[tagline]".
Motion: smooth easing, no bouncing, no particles, no lens flares. Outgoing text leaves before incoming text arrives.

IDEA: one hero item from [your product] (e.g. a cover image or a featured card) threads through the whole film: it sits in the UI, it is what the feature finds, it fills the stage in the hero moment, the floating cards orbit it, and at 14–15s the cards collapse inward into the logo.

SOUND: a mid-tempo electronic house bed around 120 BPM with clean four-on-the-floor kicks and synth bass, running without a break. A low boom and rising sweep as the logo arrives; a tight swoosh on every layout change; key clicks while anything is typed and a soft tick when a result appears; air whooshes on wipes; a pitch-rising swoosh as the cards collapse into the logo; a clean bass hit on the final lockup, with a chord decaying under the tagline. No narration.
Show full prompt ▾Show less ▴

Try this prompt on iart.ai →

When to use it. Best when you know the exact story and brand look. Not for exploring ideas; the model does what you wrote and adds little of its own.

4. The classic SaaS launch

The most copied prompt in the collection: at least six creators posted it nearly word for word. It asks for "these typical, very professionally edited, motion-graphics-styled product launch videos that you see people making on Twitter." The result is the familiar format: problem, logo, three features shown in the UI, CTA.

The prompt (adapted from @moritzkremb's original):

I want you to create a highly professional SaaS product launch video for [your product]. Use my attached screenshots as the reference for the UI. Turn it into these typical, very professionally edited, motion-graphics-styled product launch videos that you see people making on Twitter when they launch new SaaS products (showing off the features, the benefits, and all of these things).

IDEA: the mess gets sucked into one place. Scattered files collapse into your logo, then a cursor tours [your product]'s real screens doing real actions feature by feature, then pulls out to show everything.

0–3s: pure white. Untidy file cards named like [the files and tools your users juggle] (e.g. "Plan_FINAL_v3", "Untitled (4)") drift inward. Centred bold text swaps fast: "[pain 1]" → "[pain 2]" → "[pain 3]?". At 2.75s the cards swirl and collapse into a point.
3–6.5s: the point bursts into your logo, the wordmark and "[tagline]". Brief hold.
6.5–12.6s: dissolve into [your product]'s main screen (sidebar left). Pill tag "[feature 1]", headline "[its benefit]". The cursor does the key action (e.g. a command menu; three items appear). Headline swaps to "[the result]"; the cursor ticks the items off in quick succession, then clicks the next page in the sidebar.
12.6–18.6s: pill "[feature 2]", headline "[its benefit]". The cursor switches view; rows reorganise fluidly into columns; it drags one card to another column, where it snaps in.
18.6–23.9s: pill "[your standout feature]". A prompt bar slides up and types a request; headline "[the outcome]"; a structured result streams in line by line; the cursor clicks to confirm.
23.9–28s: zoom out into a tilted grid of dozens of [your product]'s screens or templates. Big bold words cut centre, one per beat: "[A]." "[B]." "[C]." then "[all-in-one line]".
28–32s: fade to white. Logo drops in, wordmark, subhead, a black pill button "[call to action]" over [your website]. A cursor enters lower right and clicks it. Hold.

LOOK: pure white, light-grey borders and card surfaces, black type; [your product]'s UI colours only as small highlights (checks, status tags). Neo-grotesque sans, bold sentence-case headlines. Centred, UI faithful to the screenshots.

MOTION: snappy demo tempo. Headlines slide vertically with a quick damped overshoot; the UI reacts instantly like the real app; smooth camera pull-outs, no motion blur.

SOUND: upbeat electronic around 115–120 BPM with marimba arpeggios and crisp hats, dipping under effects. Suction whoosh on the collapse; a pop as the logo lands; key clicks and one click per checkbox; a swoosh on the view switch, a soft thud as the card drops; a shimmer as the result streams; a riser under the zoom-out with a bass hit per word; a button click and final chord. No narration.
Show full prompt ▾Show less ▴

Try this prompt on iart.ai →

When to use it. Best for a feature-led SaaS launch that needs to look like the genre. Not for standing out; everyone else used the same prompt.

5. The one-shape UI loop

One rounded shape is never cut. It becomes the button, then the form, then the confirmation, changing size, corner radius and color while a cursor clicks and types through the real flow. The last frame matches the first, so the video loops. This is the Dribbble-style UI motion you see on landing pages.

The prompt (adapted from @verbove's original):

Make a looping 16:9 product video for [your product] using my attached screenshots as the reference for the UI.
Rules: one shape, never cut. Every state is the same element changing size, radius and color while the content swaps. A cursor drives the sequence with real clicks, typing and one drag. Real UI, real data, no placeholders.
Structure: 120 BPM, something happens on every beat: logo → [state 1] → [state 2] → [state 3] → … → logo.
Motion: springs with only a tiny overshoot; content enters after its container starts morphing and leaves before the next morph; zoom so every state fills the frame; the last frame equals the first.

IDEA: one centred shape carries one complete task in [your product], growing from a logo pill into every screen and back.

0–2s: a bright accent pill holds your logo and wordmark. A cursor glides in; content blurs into "[main action] ↗".
2–4s: click (the pill dips and eases back); it expands into a large white rounded card.
4–8s: [state 1, the main input]: the cursor types a number digit by digit; a calculated result updates in lockstep; click "Continue".
8–11s: [state 2, choosing something]: three letters typed, the list filters live, one row is clicked.
11–14.5s: [state 3, review]: line items drop in one by one; click confirm.
14.5–16.2s: the card shrinks into an accent square, then a circle; a progress ring traces once around your logo mark.
16.2–19.5s: a tall card: four timeline steps tick green in turn; a badge flips "[in progress]" → "[done]".
19.5–21.2s: a compact accent success card; a check pops, [one key number from your product] rolls up from zero; click "Done".
21.2–25s: wide overview: a total rolls up, a row of cards slides in, click a tab.
25–28s: [one more feature]: an object in perspective (e.g. a card) flips in 3D on one tab click and back on the other.
28–30s: everything collapses, corners rounding, into the starting pill; the cursor leaves.

LOOK: your brand colours in two roles: a deep vignetted dark tone for the stage, a bright accent for the pill and success states; white cards, charcoal text, grey labels. Bold grotesque, heavy numbers. Big radii, soft shadows, strictly centred.

MOTION: brisk, with half-second micro-holds for reading. Morphs snap then decelerate long; counters roll fast and land softly.

SOUND: a minimal electronic percussion bed with sub swells, never silent. A key click per character, a click on every button, a ticking roll under counters, an air whoosh on each morph, a charging riser into the ring, an ascending chime per timeline step and a completion ping, a swoosh per flip, a suction into the closing pill. No narration.
Show full prompt ▾Show less ▴

Try this prompt on iart.ai →

When to use it. Best for apps with a clear step-by-step flow (checkout, onboarding, sending money). Not for products whose value isn't visible in one flow.

6. The emotional story

It starts from nine words: "build me an incredible video for my app. make it emotional." The model invents a person, a problem and a payoff your product makes possible. Ours wrote a grandmother's Spanish birthday letters, a year-long streak, and a last line that turns into a phone call.

The prompt (adapted from @marklaunches's original):

Build me an incredible video for my app [your product] ([one line on what it does]). Make it emotional.

IDEA: a small personal story around one object: [something from a person the user loves that they couldn't fully use or understand, e.g. a handwritten card in a language they can't read]. A fast montage of daily use as a streak counter climbs; then the object, unlocked, and one moment with that person.

0–6.5s: dark warm-brown vignette; the object centred (e.g. a cream card where ink handwriting writes on word by word). Subtitles: "[Every year, she did this for me.]" then "[I never understood it.]"
6.5–13s: cut to a phone centred on a dark stage in your brand colour: [your product]'s first small task (e.g. three tiles fly into place), a success banner. Big bold "DAY 1" settles at the right. Subtitle: "[I started small.]"
13–15s: morning peach background; next task; a streak counter at the right rolls 7 → 30; a reminder notification drops in. Caption: "[Before work.]"
15–17s: sky blue; 30 → 100; "[On the bus.]"
17–19s: night navy with stars, clock near midnight; 100 → 200; a playful notification; "[At midnight.]"
19–23s: dusk purple; the counter accelerates to [your big number] with a flare; "[On days I didn't want to.]" then "[365 days.]"
23–33s: back to the object, slow push in. Under each handwritten line a small green line wipes in with a thin underline: what [your product] unlocked (e.g. the translation). Subtitle: "[Then, one evening, I read it again.]"
33–39s: fade to black. A call screen: round contact icon with an initial, "[their name] ♥", "calling...", then a timer ticking 00:00–00:03 and a pulsing green waveform. Two subtitle lines, green italics over white translation: "[what I finally say]" / "[their surprised reply]".
39–44s: white end card: your logo or character drops in with a bounce, then the wordmark, "[tagline]", "[where to get it · your website]".

LOOK: warm espresso brown, cream paper and navy ink for the object; your brand colour, then one colour field per time of day, for the montage; black call screen; white end card. Casual handwriting for the object, a rounded geometric sans for subtitles and UI. Centred, counter right, subtitles bottom centre.

MOTION: slow, quiet bookends around a fast montage. Springy taps, banners and counters in the app; clean cuts, colour morphs through the montage, a gentle fade into the call.

SOUND: tender fingerpicked guitar throughout. Pen scratching on the write-on; tile taps and a bright success chime per task; a rolling tick for each counter run; a ping per notification; a soft marker whoosh per translated line; a call connect beep; a final guitar chord and short chime on the end card. No narration.
Show full prompt ▾Show less ▴

Try this prompt on iart.ai →

When to use it. Best for consumer apps where the product changes someone's life in a way you can show. Not for B2B tools; a heartfelt story about a spreadsheet reads as parody.

How to make your own launch video on iArt

  1. Open iart.ai and choose Animation.
  2. Paste one of the prompts above and replace the brackets with your product.
  3. Optional: attach screenshots or your logo with the + button, or paste your site's address. They appear in the film as they are.
  4. Send. Opus 5.5 plans the film, writes it and shows it in the preview.
  5. Refine by chatting ("make the logo reveal faster"); for narration, choose a voice in the message box. Then export the MP4 (paid plans).

Prefer a dedicated workflow? The product launch video maker is built for exactly this — brief in, launch film out.

FAQ

Can Opus 5.5 make a product launch video from one prompt?

Yes. Every film in this article came from one prompt, plus at most one follow-up message. The model draws every frame in code, so it suits motion graphics, UI and type; it doesn't film a physical product.

How long should a launch video prompt be?

One or two sentences are enough for a good launch film: what it is, how long it runs, the format and the moment that matters. For a more professional film, add a short brief of about 300 to 450 words: the one idea that carries the film, the shots second by second with the exact on-screen words, the look, the motion, and a sound for each action. A brief takes longer to make and uses more credits.

Do I need to know how to code?

No. In iArt's Animation mode you type the prompt, Opus 5.5 writes and checks the code, and you watch the preview and, on a paid plan, export an MP4.

Can I use my own screenshots and a voiceover?

Yes. Attached screenshots and logos appear in the film as they are, not redrawn, and you can give it your site's address instead. For a voiceover, choose a voice in the message box.

How long should a launch video be?

15 to 20 seconds for social posts. The launch videos that spread after the Opus 5.5 release were almost all in that range, with something new on screen every second. Go to 30 seconds only if you need to explain how the product works. Wistia's 2026 State of Video, drawn from over 13 million videos, found the same pattern: the shorter the video, the higher the engagement rate (Wistia).

What does it cost?

New iArt accounts start with free credits, enough to try a first short film. See the pricing page for plans.

Can I reuse the prompts?

Yes. Each prompt is adapted from a creator's public post, credited and linked in its section. Credit them if you share their wording.

Pick the prompt that matches the video you'd post, swap in your product, and make your launch video on iart.ai.

Ready to create stunning motion graphics?

Turn your ideas into professional animations with AI