10 AI Video Prompt Examples That Actually Work

AI video prompt examples work best when each one names a subject, an action, a setting, a camera move and the light, plus sound on models that render audio. Below are 10 copy-paste prompts for cinematic, product, pet, food and abstract shots, each built from those parts, so you can swap the subject and paste them into any text-to-video generator.
Most disappointing AI videos come from prompts like "a cool city at night". The model has to guess the subject, the camera, the light, and the sound — and four guesses rarely all land. The fix is structural, not creative genius.
Every prompt below follows the same six-part structure we use in the full text-to-video guide:
Subject + action + setting + camera + light (+ sound, if supported).
Paste any of these into the AI video generator, tweak the subject, and you're most of the way there.
Choose the model before relying on the sound direction. Seedance 1.0 Pro Fast
and Hailuo 2.3 produce silent clips; a Sound: line does not add audio to them.
Use the AI Model Finder to compare the model, supported
settings and credit estimate for your scene. Hailuo 2.3 also requires a source
photo, so it cannot make these scenes from text alone.
Cinematic & landscape
1. Coastal drone shot
Aerial drone shot flying over a turquoise coastline at golden hour, waves crashing on white cliffs. The camera glides forward, then tilts down over a sailboat. Sound: cinematic orchestral swell, ocean waves, wind.
2. Rainy street mood piece
A narrow old-town street at night in heavy rain, neon signs reflecting in puddles. Slow dolly forward at street level, shallow depth of field. Cool blue-and-pink light. Sound: rain on pavement, distant thunder, soft synth pad.
3. Mountain time-lapse feel
Clouds rolling over a jagged mountain ridge at sunrise, light sweeping across the valley. Fixed wide camera. Golden light breaking through mist. Sound: low ambient drone, wind gusts.
Scenic shots like these double as B-roll for tourism and hospitality marketing — pair them with a talking host from the travel video generator when the clip needs a narrator.
Product & commercial
4. Luxury product pedestal
A sleek perfume bottle on a black marble pedestal, rotating slowly as golden light sweeps across the glass. Water droplets splash around it in slow motion. Studio lighting, luxury commercial style. Sound: elegant piano, soft whoosh.
5. Sneaker drop
A white sneaker falling in slow motion onto wet asphalt, splash radiating outward, camera orbiting 180 degrees around the impact. Dramatic rim lighting. Sound: deep bass hit, slow-motion whoosh, rain ambience.
For ads built around a product photo you already have, the product video ad flow is usually the better tool — it builds the ad from your actual product photo instead of inventing one.
Character & animal
6. Cute pet clip
A fluffy orange kitten chases a butterfly across a sunlit kitchen counter, slides on the smooth surface, and tumbles into a soft towel. Warm morning light, shallow depth of field. Sound: playful ukulele, tiny meows.
7. Sci-fi walk
A lone astronaut walks across a red desert planet toward a massive derelict spaceship half-buried in sand. Dust storms swirl in the distance, twin suns setting. The camera slowly pushes in from behind. Sound: deep ambient drones, wind, heavy breathing.
Food
8. Ramen close-up
Extreme close-up of steaming ramen — chopsticks lift glossy noodles as broth drips in slow motion, chili oil swirling on the surface. Neon restaurant lights bokeh in the background. Sound: gentle sizzle, cozy lo-fi beat.
9. Coffee pour
Espresso pouring into a glass cup in macro slow motion, crema folding over itself, steam curling upward. Fixed camera, dark moody kitchen light. Sound: liquid pour, soft café ambience.
Food shots like these are B-roll; if you want the owner talking over the espresso pour, the cafe video generator handles the talking-head half — and the menu special video maker covers weekly promos.
Abstract / background loops
10. Ink in water
Colored ink blooming underwater in slow motion — violet and teal clouds unfolding against a black background. Fixed macro camera. Sound: deep ambient tone, soft underwater movement.
Test a prompt right now
Choose a model, resolution and length, then review the credit estimate before rendering. One-time credit packs start at $5.
Open the AI video generatorWhy these work — the checklist
Before you render, check the five visual parts and decide whether you need sound:
- Subject — one clear thing the shot is about.
- Action — a verb; motion is what video is for.
- Setting — place the subject in a specific environment or background.
- Camera — name the move (track, orbit, push-in, fixed) or the model invents one.
- Light — "golden hour", "neon", "studio" — one phrase sets the grade.
- Sound — on a model with native audio, describe the music, ambience or dialogue you want. For a silent model, plan to add audio separately.
Test a short clip at a lower resolution supported by your chosen model, then increase quality once the motion works. 480p and 1080p are not available on every model. The full guide explains model-specific controls and a qualified cost example.
How do you write a prompt for an AI video generator?
Write it like a one-shot shot list: who or what is on screen, what it does, where it happens, how the camera moves and what the light looks like. Keep one main action per clip, because a single short video rarely handles several events well.
Start from one of the sample prompts above, change only the subject, and render a short test. If the motion is wrong, rewrite the action or camera phrase rather than adding more adjectives. When you are ready to try your own, paste it into the AI video generator and check the credit estimate before rendering.
How do you write image-to-video prompts?
In image-to-video, the photo already sets the subject, colors and composition, so the prompt should describe motion instead of appearance. Say what moves, how the camera moves and how the shot ends, and avoid re-describing details that are already in the picture.
For example, with a product photo: "The bottle rotates slowly on its pedestal while soft light sweeps across the glass. Slow push-in, fixed background." Models that need a start image, such as Hailuo 2.3, work this way. If the photo is of a product you sell, the product ad maker builds the ad from that photo for you.
How long should an AI video prompt be?
Two to four sentences is usually enough, about the length of the examples on this page. Longer prompts tend to pack in competing actions and details that the model drops or blends together.
Put the subject and action first, then the setting, camera and light, and add a short sound line only on a model that renders audio. If a result comes out generic, make the existing words more specific before you make the prompt longer.


