- What an AI video prompt is in 2026
- The tools these prompts are written for
- A worked example: turning ten prompts into a month of content
- How to turn any prompt into a usable clip, step by step
- 50 copy and paste AI video prompts for short clips
- Why one prompt rarely gives you one usable clip
- Frequently asked questions
- Primary sources
The short version: These 50 prompts are written for tools like Sora, Runway, Kling and Luma Dream Machine, and every one is built to be copied, pasted, and edited with your own product, colour and setting swapped in. Copy paste prompts save you the fifteen minutes most people waste typing and deleting before they get anything usable, but they will not fix a model having an off day, so budget time to run each prompt two or three times before you pick your favourite.
What an AI video prompt is in 2026
An AI video prompt is a short written description that a text-to-video model turns into a clip, usually between three and ten seconds. The difference between a prompt that works and one that doesn't is rarely the idea. It's the detail. You need a subject, an action, a camera move, a lighting description, and a duration. Leave any of those out and the model fills the gap with a guess, and its guesses are often odd hands, warped text, or a shot that drifts for no reason.
That's why every prompt in this list follows the same pattern: subject, camera direction, light, length. Swap the bracketed words for your own product or setting and the structure does the heavy lifting.
The tools these prompts are written for
You don't need five subscriptions. Pick one and get good at its quirks. As of early 2026, the landscape looks roughly like this:
- Sora, built into ChatGPT Plus at around $20 a month in the US and roughly the same in pounds in the UK, with clip lengths generally capped around 10 to 20 seconds depending on your tier.
- Runway, with a Standard plan around $12 a month billed annually, giving you a set number of monthly credits and clips typically limited to about 10 seconds before you need to extend them.
- Luma Dream Machine, which offers a free tier of a limited number of generations a month, with paid plans starting around $9.99 a month for more credits and faster processing.
- Kling AI, which runs on a credit system through its app and web tool, with longer clip lengths available on higher tiers.
- Google's Veo, increasingly bundled into higher Gemini subscription tiers rather than sold as a standalone product.
All of them watermark free or lower-tier outputs, so if you're building anything for a paying client, factor the upgrade cost in before you promise a watermark-free deliverable.
A worked example: turning ten prompts into a month of content
Say you run a small candle business and you need ten short clips for Instagram Reels this month, on a budget of maybe an hour a week. You'd pick three or four product prompts from the list below (the macro close-up, the pour shot, the unboxing shot), two atmosphere prompts (the flickering candle flame, the misty morning shot), and one or two lifestyle prompts showing hands wrapping a gift. That's six prompts. Run each one twice, which gives you twelve raw clips. Keep the best eight, trim them to three or four seconds each in CapCut, add a text overlay with your product name, and you've got a week of content from about ninety minutes of work spread over two sittings. The maths only works because you generated more than you needed and threw half of it away without guilt.
How to turn any prompt into a usable clip, step by step
- Write the prompt with all four elements: subject, camera move, lighting, duration. Don't skip lighting, it's doing more work than people assume.
- Generate at least three versions of the same prompt. One will usually be noticeably better than the other two.
- Trim in your editor to the strongest two to four seconds rather than using the whole clip.
- Add text or motion graphics in Canva or CapCut, since AI models still render on-screen text badly more often than not.
- Match the clip to a trending audio track or your own voiceover.
- Export vertical at 9:16 for Reels and TikTok, horizontal at 16:9 for YouTube, and keep both versions saved so you're not re-exporting later.
If you're producing these for a client rather than yourself, it's worth writing the content brief before you touch a prompt box, the same way you'd approach a structured blog brief. Knowing the platform, the audience and the call to action before you generate anything saves you from a folder full of pretty clips that don't fit anywhere.
50 copy and paste AI video prompts for short clips
Product close-ups and b-roll
- Macro close-up shot, slow dolly in, of [your product] resting on a linen surface, soft window light from the left, shallow depth of field, 5 seconds, no text.
- Top-down flat lay, camera slowly rotating 15 degrees, of [your product] surrounded by [three related props], warm natural light, 6 seconds.
- Slow motion shot of [liquid product] being poured into a glass vessel, studio lighting, dark background, 4 seconds.
- Handheld tracking shot following [your product] as it is picked up off a wooden shelf, natural daylight, soft focus background, 5 seconds.
- 360 degree rotating product shot of [your product] on a plain white cyclorama background, even studio lighting, no shadows, 6 seconds.
- Extreme close-up on the texture of [your product], camera panning left to right, soft diffused light, 4 seconds.
- Slow motion unboxing shot, hands opening a box to reveal [your product], warm tungsten lighting, cosy home setting, 6 seconds.
- Overhead shot of [your product] being placed into a gift bag surrounded by tissue paper, bright natural light, 5 seconds.
Faceless brand and behind-the-scenes clips
- Static shot of a laptop screen showing [your dashboard or app], soft office lighting, hand typing on keyboard, 6 seconds, no visible face.
- Over-the-shoulder shot of hands writing in a notebook at a desk, coffee cup beside it, morning light through a window, 5 seconds.
- Slow pan across a tidy desk setup with laptop, plant and notebook, warm afternoon light, subtle camera shake for realism, 6 seconds.
- Close-up of hands sorting through printed documents on a table, natural light, shallow depth of field, 5 seconds.
- Static wide shot of an empty modern office space, morning light streaming through blinds, dust particles visible, 6 seconds.
- Slow zoom on a whiteboard covered in sticky notes, a hand circling one note with a marker, bright office lighting, 5 seconds.
- Tracking shot following a pair of hands scrolling through a phone showing [your app or website], soft light, blurred background, 4 seconds.
- Close-up of a hand placing a business card on a desk next to a cup of coffee, natural window light, 4 seconds.
Nature and atmosphere backgrounds
- Wide aerial-style shot of rolling green hills at golden hour, slow push forward, warm colour grade, 6 seconds.
- Slow motion shot of waves rolling onto a sandy beach at sunset, static camera, orange and pink sky, 6 seconds.
- Close-up of rain falling on a window pane with a blurred city skyline behind it, moody grey lighting, 5 seconds.
- Time-lapse style shot of clouds moving across a blue sky above a forest, natural colour grade, 5 seconds.
- Static shot of a candle flame flickering in a dark room, warm orange light, shallow depth of field, 4 seconds.
- Slow pan across a misty forest at dawn, soft blue light, birds faintly visible in the distance, 6 seconds.
- Close-up of autumn leaves falling in slow motion, soft backlight, shallow depth of field, 5 seconds.
- Wide shot of snow falling gently over a quiet street at night, warm streetlights glowing, 6 seconds.
Food and drink
- Slow motion pour of coffee into a white ceramic mug, steam rising, warm kitchen lighting, 5 seconds.
- Close-up of a knife slicing through a fresh loaf of bread on a wooden board, natural light, 4 seconds.
- Overhead shot of hands plating a dish with a garnish, warm restaurant lighting, shallow depth of field, 5 seconds.
- Slow motion shot of a drink being stirred with ice clinking against the glass, dim bar lighting, 4 seconds.
- Close-up of butter melting on a stack of pancakes, warm morning light, steam visible, 5 seconds.
- Static shot of a bakery display case with pastries, soft golden lighting, shallow depth of field, 5 seconds.
- Slow motion shot of sauce being drizzled over a dessert, dark background, studio lighting, 4 seconds.
Lifestyle and social proof style clips
- Handheld shot of someone typing on a laptop at a coffee shop table, natural window light, blurred background, no visible face, 5 seconds.
- Close-up of hands writing a five star review on a phone screen, soft light, shallow depth of field, 4 seconds.
- Static shot of a delivery box being carried up to a front door, natural daylight, 5 seconds.
- Slow motion shot of hands shaking after signing a contract on a desk, warm office light, 4 seconds.
- Wide shot of a small team gathered around a laptop in a bright office, natural light, subtle camera movement, 6 seconds.
- Close-up of a hand highlighting text in a printed report with a yellow marker, natural light, 4 seconds.
- Tracking shot of someone walking into a shop front with a warm glowing sign, evening light, 5 seconds.
Motion graphics and text-driven promos
- Clean animated text reveal, bold sans-serif font, words "[your headline]" sliding in from the left on a solid colour background, 3 seconds.
- Animated countdown numbers, 3, 2, 1, in bold white text on a black background, subtle glow effect, 3 seconds.
- Simple animated arrow pointing towards a product placeholder, minimal flat design, pastel background, 3 seconds.
- Text overlay animation reading "[percentage]% off" bouncing gently into frame, bright colour background, 3 seconds.
- Animated progress bar filling from 0 to 100 percent, clean flat design, brand colour background, 3 seconds.
- Kinetic typography of the words "[your call to action]" appearing letter by letter, bold font, dark background, 4 seconds.
Seasonal and trend-led clips
- Close-up of hands wrapping a Christmas present with red ribbon, warm fairy lights blurred in the background, 5 seconds.
- Slow motion shot of confetti falling in front of a plain background for New Year content, colourful lighting, 3 seconds.
- Static shot of a jack-o-lantern glowing on a porch step at dusk, warm orange light, 5 seconds. If you're building out a wider seasonal push, it's worth pairing this with something like Halloween-themed content ideas so the visuals and the copy match.
- Close-up of hands arranging eggs in a woven basket for spring content, soft natural light, 4 seconds.
- Slow pan across a table setting with candles and roses for a romantic occasion, warm lighting, 5 seconds.
- Wide shot of fireworks exploding over a night skyline, static camera, 5 seconds.
Why one prompt rarely gives you one usable clip
There's a version of this topic that gets sold as effortless, and I want to be straight with you about the bit that isn't. On any given prompt, across every one of these tools, you should expect roughly one in every three to five generations to be usable. The other two to four will have a warped hand, a camera move that drifts oddly, or a background object that flickers between frames. That's not a sign you wrote a bad prompt. It's the current state of the technology, and treating it as a one-shot tool is the single biggest reason people give up after their first attempt.
The second thing nobody likes to say out loud is that character consistency across multiple clips still doesn't work reliably outside a handful of paid features designed specifically for it. If you're picturing a recurring "presenter" appearing in clip after clip with the same face and outfit, most of these tools will give you a different-looking person every time unless you're on a plan built for that. Plan your content as standalone b-roll rather than a continuous story, and you'll stop fighting the tool.
There's also a practical use case worth naming directly. If you sell property, these same prompt structures work well for atmospheric shots to sit alongside your listing copy, similar to how you'd handle property description writing, a slow pan across a hallway or a close-up of a front door catching evening light can carry a listing further than another static photo. And if you're repurposing a clip into a paid ad, it's worth writing the ad copy at the same time rather than as an afterthought, the same discipline you'd apply to Facebook ad copy, because a beautiful clip with a flat caption still won't convert.
Want AI doing the heavy lifting in your marketing?
I build the systems that handle the boring 80 percent, so you get your week back. Done properly, with the human kept in.
One more honest note on ethics and rights. Don't generate clips that mimic a real, identifiable person without consent, even in a "faceless brand" context, and don't lift a customer's actual words into a video script without checking with them first, the same care you'd take turning a real testimonial into a written case study. The prompts above are all built around anonymous hands, objects, and landscapes precisely to sidestep that problem.
Related guides live in the ChatGPT Prompts and Limits: 30 Guides to Fun Prompts, Photo Edits and Daily Caps. To write your own, try the free ChatGPT prompt generator.
Frequently asked questions
Which AI tool is best for short video clips in 2026?
There's no single best option, only a best fit for your budget and platform. Sora suits people already paying for ChatGPT Plus who want video without a second subscription, Runway suits people who want more editing control over the output, and Luma Dream Machine suits anyone testing the water on a free tier before committing to a paid plan.
Can I use these prompts for commercial content?
Generally yes, but check the specific tool's terms before using a clip in paid advertising, since some platforms restrict commercial use on their free tiers or require attribution. Read the terms of service for whichever tool generated the clip rather than assuming all AI video tools license their output the same way.
Why do my AI video clips look different every time I use the same prompt?
Text-to-video models introduce randomness into every generation by design, so identical prompts produce different results each time. This is normal, not a fault, and it's why generating three or more versions of each prompt before picking a winner is part of the workflow rather than an optional extra.
How long should an AI-generated short clip be?
Most usable output sits between three and six seconds once you've trimmed the raw generation, even if the tool technically produces longer clips. Short, clean segments cut together in an editor almost always look better than one long AI-generated shot used untouched.