AI Video Prompts for Products: A Category-by-Category Prompt Library
Different product categories break AI video in different ways — fabric moves differently than liquid, and a face forgives nothing. The fix is category-specific base prompts: 20 templates below across fashion, beauty, tech, home, and food, each with slots you swap per SKU. Two rules carry across all five categories: prompt one physical truth per shot (drape, viscosity, reflection, weight, steam), and anchor exact product looks with image-to-video from a real packshot. Everything else — camera, light, length — follows the category cheat sheet in each section.
Why Category-Tuned Prompts Matter in 2026
Generic prompts produce generic renders. The commerce numbers are not generic: 82% of consumers have purchased after watching a product video (webtonic) and 73% prefer learning about products through short video (sci-tech-today). But the clip only converts if the material behavior looks true — knits must drape, creams must shear, food must steam. Category prompts encode those physics.
Full disclosure: VideoIdeas builds AI-ready prompts from live social trends with category and product slots prefilled. The library below is the manual version — see how it works.
How to Adapt Any Base Prompt
Swap four slots and one base prompt serves a whole category:
| Slot | What to write | Bad example | Good example |
|---|---|---|---|
| [product] | Exact item + finish | “our jacket” | “cropped puffer jacket, matte olive shell” |
| [material truth] | The physics it must show | “looks nice” | “fabric drapes and compresses at the sleeve” |
| [context] | Where it lives | “outside” | “wet city street at dusk” |
| [action] | One motion verb | “shown in use” | “hand lifts the collar and releases it” |
Add fixed constraints to every prompt in the set: “single continuous shot, [6] seconds, [9:16], no text, no logos, realistic materials.”
Fashion Prompts (1–4)
Fashion fails on fit and drape. Motion must come from the garment or a hand — never both and the camera at once.
1. Fabric drape hero
[product: garment + color + cut] hanging on a minimal rack, [material truth: fabric drapes
and sways gently in a soft breeze]. Camera: slow pan top to bottom, 50mm. Light: diffused
daylight from [direction], [setting: studio loft]. Mood: [tone], editorial calm.
Constraints: 6s, 9:16, one take, garment shape constant, no text.
Use it for: PDP hero clips for dresses, shirts, and outerwear. · Best tool: image-to-video from a hanging product shot.
2. On-body fit loop
[persona: body type + styling] turning slowly in [context: fitting room / street corner]
wearing [product], camera static mid-shot. The garment settles naturally with movement —
[material truth: waistband stays put, hem swings]. Natural light, realistic proportions,
7 seconds, no text.
Use it for: fit questions that photos can’t answer. · Best tool: UGC ad generators or image-to-video with a model plate.
3. Detail-and-craft macro
Extreme macro creeping along [construction detail: stitching / zipper teeth / knit rib]
of [product]. Raking side light reveals texture depth, shallow focus. Warm [tone].
6 seconds, one continuous creep, realistic fiber detail.
Use it for: premium and handmade positioning. · Best tool: image-to-video.
4. Styled-look transition
Same [persona] and framing in two beats: [look A styling] shifts to [look B styling]
with [product] as the constant piece, via a single fabric-sweep transition.
Locked camera, even daylight, plain background. 6 seconds total.
Use it for: “one piece, three ways” content. · Best tool: two image-to-video renders composited in post.
Beauty Prompts (5–8)
Beauty fails on skin and hands. Favor macro product textures and indirect application moments over full faces.
5. Texture swatch macro
Macro close-up: [product: shade + finish] swatched on [surface: forearm / glass tile],
[material truth: cream shears and settles to a satin finish / powder kicks and falls].
Hard directional light, high-speed look, dark neutral background. 5 seconds, no text.
Use it for: shade and finish proof — the single highest-need beauty asset. · Best tool: text-to-video or image-to-video.
6. Dropper and viscosity
Close-up of a glass dropper lifting [product: serum / oil] over the open bottle,
[material truth: single viscous drop forms and falls slowly]. Backlit, warm soft light,
marble surface below. Luxury still-life mood, 5 seconds, realistic fluid physics.
Use it for: serum and oil hero clips. · Best tool: text-to-video.
7. Application ritual (hands only)
Hands-only mid-shot: fingers press [product] into [skin area or tool] in one slow motion
over [surface: vanity]. Soft morning window light, natural skin texture, no face visible.
Calm ritual mood, vertical 9:16, 7 seconds.
Use it for: routine content without model-face risk. · Best tool: UGC ad generators.
8. Shelfie rotation
[product family] arranged on a [pastel / stone] shelf, camera slides slowly along the
lineup and settles on [hero SKU]. Soft bounced light, gentle shadows, no reflections of
gear. 7 seconds, no printed labels visible.
Use it for: range ads and gifting sets. · Best tool: image-to-video.
Tech and Gadget Prompts (9–12)
Tech fails on geometry and reflections. Lock both with constraints.
9. Precision edge light
[product: device + finish] on a dark [surface], a thin light band travels along its edge
highlighting [design line / chamfer]. Camera static, 85mm, deep blacks, cool rim light.
Engineered mood, 5 seconds, no screen text, device geometry constant.
Use it for: premium hardware positioning. · Best tool: image-to-video.
10. Hands-scale interaction
First-person hands pick up [product] and perform [one interaction: press a button /
fold the hinge / plug in a cable] over a clean desk. Even task lighting, slight handheld
motion, realistic screen glow. 6 seconds, all ports and buttons anatomically correct.
Use it for: usability proof and PDP embeds. · Best tool: image-to-video.
11. Exploded components
[product] at center, its components — [component list] — separating gently apart in
mid-air and settling back together, on a seamless [gray] backdrop. Technical studio light,
precise mechanical motion, camera slowly orbits 20 degrees. 8 seconds, no labels.
Use it for: build-quality claims and launch teasers. · Best tool: multi-shot text-to-video.
12. Desk-setup placement
[product] placed in a lived-in [desk / kitchen] setup among [props: notebook, mug, plant],
a hand enters and uses it once. Warm practical lamp light, slow camera drift left to right,
realistic reflections on [surface]. 7 seconds.
Use it for: lifestyle fit for work-from-home audiences. · Best tool: text-to-video.
Home and Living Prompts (13–16)
Home fails on scale and light. Always include a reference object and a time of day.
13. Room-corner reveal
Camera slides past a doorframe to reveal [product] in a [room: sunlit bedroom / reading
nook], a hand adjusting it once. Late-afternoon sun, dust in the light beam, soft shadows.
Slow reveal, 7 seconds, realistic scale versus [reference object].
Use it for: decor and furniture prospecting. · Best tool: text-to-video.
14. Material-and-softness proof
Close-up: a hand presses into [product: cushion / mattress / towel] and releases —
[material truth: fabric rebounds slowly, weave compresses and recovers]. Warm side light,
texture-forward framing, [surface] beneath. 5 seconds.
Use it for: comfort claims where feel is the sell. · Best tool: text-to-video.
15. Function demo loop
Hands demonstrating [product function: fold / adjust / assemble] on [surface], one
continuous action, camera static mid-shot. Even daylight from a window, honest home
lighting, [reference object] in frame for scale. 7 seconds.
Use it for: assembly and utility proof that cuts returns. · Best tool: UGC ad generators.
16. Golden-hour ambiance
[product] in a [room] at golden hour, slow light shift across [surface] as the sun moves,
long soft shadows. Camera on a gentle dolly-in. Warm [tone], lived-in but tidy. 8 seconds,
no text.
Use it for: mood-led brand ads and seasonal sets. · Best tool: text-to-video.
Food and Beverage Prompts (17–20)
Food fails on appetite physics. Steam, pour, crumb, and condensation must be explicitly prompted.
17. Steam and pour hero
Close-up: [beverage] poured into [vessel] beside [product], steam rising and curling
in backlight. Slow motion, hard backlight through the steam, dark warm background.
[material truth: liquid viscosity realistic, pour steady]. 5 seconds, no text.
Use it for: coffee, tea, and beverage brands. · Best tool: text-to-video.
18. Texture bite macro
Macro: [food item] breaking / biting open to reveal [interior: flaky layers / melted
center / crumb structure]. High-speed camera look, bright appetizing key light, shallow
focus. 4 seconds, realistic crumb and steam.
Use it for: snack and bakery feeds. · Best tool: text-to-video.
19. Ingredient assembly
Top-down: [ingredients] dropping one by one into [bowl / blender] around [product],
each landing with realistic weight and bounce. Even overhead daylight, [surface: wood
board]. Playful rhythmic pacing. 7 seconds.
Use it for: recipe-style ads and meal-kit content. · Best tool: multi-shot text-to-video.
20. Cold-chain freshness
Macro: condensation beading and rolling down [chilled product / can / bottle] against a
[background], [material truth: droplets form, merge, and run]. Hard side light, high-speed
look, deep contrast. 5 seconds.
Use it for: refreshment positioning in warm-weather campaigns. · Best tool: text-to-video.
Fill-in-the-Blank Hook Lines by Category
Add as captions or VO — generated text stays out of the prompts.
- Fashion: “The fit detail on [garment] that photos never show: [detail].”
- Fashion: “How [fabric] actually falls on a [body type]. No filter.”
- Beauty: “[Shade] in [light condition] — the only swatch that matters.”
- Beauty: “What [product texture] does in the first 10 seconds on skin.”
- Tech: “Everyone checks [feature]. Nobody checks [build detail].”
- Tech: “[Timeframe] of daily use later — the hinge/port/battery truth.”
- Home: “It looked [adjective] in the photo. Here’s the real scale.”
- Home: “Watch what happens when I press it: [material truth].”
- Food: “Sound on for [sound cue: crack / pour / sizzle].”
- Food: “The inside is the whole point. Breaking it open now.”
- Any: “POV: you stopped settling for [category problem].”
- Any: “[Number] [units sold / days tested] later, here’s the honest read.”
Production Notes Per Category
- Fashion: shoot the packshot on a rack or invisible mannequin; flat-lays anchor poorly.
- Beauty: never prompt full faces in the first three seconds; skin artifacts kill trust instantly. Macro hands and texture carry.
- Tech: state “device geometry constant, ports and buttons correct” in every prompt — this category morphs most.
- Home: always include a scale reference; returns spike when scale reads wrong.
- Food: prompt the appetite physics explicitly (steam, crumb, viscosity); models under-render them by default.
- All: measure per category — hook rate, add-to-cart, and return rate. Short-form demos lift add-to-cart by 41% on average (tagshop), and returns drop when behavior looks true.
For category-matched angles instead of blank slots, browse example outputs or generate ideas free with VideoIdeas.
FAQ
What are the best AI video prompts for fashion products?
Fabric-drape heroes, on-body fit loops, and construction macros. Prompt the physics — “fabric drapes and sways,” “hem swings,” “knit rib compresses” — and keep camera and garment from both moving complexly in one shot.
How do I write AI prompts for beauty product videos?
Lead with texture and finish: swatch macros, dropper viscosity shots, hands-only application. Avoid full faces in early frames, and specify skin and cream material behavior explicitly. Backlit close-ups hide artifact risk and read premium.
Can the same AI prompt work for tech and home products?
Not well. Tech prompts must lock geometry, ports, and reflections; home prompts must add scale references and time-of-day light. Use the same skeleton, but swap the “material truth” and constraint lines per category.
What makes food AI video look appetizing?
Prompt the appetite physics: steam curling in backlight, crumb structure on a break, realistic pour viscosity, condensation beading. High-speed camera looks and shallow focus help — and keep “realistic food texture” as a hard constraint.
How do I adapt one prompt to my whole catalog?
Freeze four slots per SKU — product clause, material truth, context, action — and share the camera/light/constraint block across the category. Batch-render per category so output differences come from the slots, not the base.
Should I use image-to-video or text-to-video for category content?
Image-to-video anchored to a real packshot for anything where exact appearance sells (fashion, beauty, tech). Text-to-video for lifestyle sets, ambiance, and appetite shots where atmosphere leads and the product is composited or brief-visible.
Where do I find category-specific video ideas that are trending?
Category templates plus trend timing beat either alone. VideoIdeas tracks live TikTok, Instagram, YouTube, and Facebook trends and returns product video prompts with category slots filled.
Written by the VideoIdeas Editorial Team, makers of VideoIdeas — live social trends turned into product video concepts and AI-ready prompts. Generate ideas free.
Sources:
- webtonic — E-commerce video marketing statistics
- sci-tech-today — Short-form video ROI statistics
- tagshop — Short-form video stats
- Wyzowl — Video marketing statistics
Share Article
