Guides

Kling AI Prompts Explained: Prompts for Product Ad Videos

By the Prompt Shelf team · Updated November 19, 2026 · 8 min read

Kling AI prompts explained is what you search for when you have a product to sell, a Kling account open, and no time to guess your way through prompt after prompt. This guide is for Shopify and DTC brand owners, freelance UGC and ad creators, and small e-commerce teams who make short vertical ads for TikTok, Reels and Shorts.

Quick answer: Kling AI turns written scene descriptions into short video clips, so a prompt is a shooting script: subject, action, setting, camera, lighting, style, duration and what to avoid. Write those parts in order, fill in real product details, generate several versions, and keep the clips where the product looks accurate.

What kling ai prompts explained actually covers

A Kling AI prompt is a written shooting script that tells the model the subject, action, setting, camera, lighting, style, duration and the things it must not do.

Kling is a text-to-video and image-to-video tool, in the same category as Sora, Veo and Runway. A prompt that reads like a wish, such as "make a cool ad for my bottle," gives you one random clip. A prompt that reads like a shot list gives you a clip you can cut into an ad. For a DTC brand, that difference decides whether AI video is a cheap way to test hooks or a cheap way to make junk.

The structure below works in Kling and in most other tools. Only clip length, audio and lip-sync differ, so a tool without built-in speech needs the spoken lines added later as voiceover.

Step by step

  1. Write the angle brief first. Open ChatGPT, Claude or Gemini and paste in the Product-to-Angle Brief prompt. Fill in your product, price and offer, target buyer, the problem in the buyer's words, the proof you can actually show, and the platform. Tell the model to ask you when a fact is missing instead of inventing one.
  2. Pick the two angles with the lowest overclaim risk. A distinct angle is a different reason to buy, not the same idea reworded. Test the two with the clearest visual proof first.
  3. Draft the script shot by shot. Use the hook prompt to write ten openers for your chosen angle, then keep the two that show something instead of only saying something.
  4. Run the claims and disclosure check before you generate anything. Delete any line that promises a result you cannot support, and note where you must label the ad or disclose AI content under your platform's rules.
  5. Paste each shot prompt into Kling and generate two to four versions. Keep the clips where the product looks accurate and discard the ones with warped shapes, extra parts or invented labels.
  6. Edit the clips together in CapCut, Premiere or your usual editor. Add your real price, code and logo as text, label the ad as required, and test variants with the A/B prompt.
Prompt part What it controls Example
Subject What is on screen Hands using a matte black insulated bottle
Action What happens, in order Presses the lid button twice; the lid pops open
Setting Where it happens Bright kitchen counter, blurred background
Camera Framing and movement Close-up from 45 degrees above, slow push-in
Lighting Mood and texture Soft morning window light from the left
Style How real it looks Realistic UGC, natural colors, no cinematic grading
Duration Clip length 5-8 seconds, vertical 9:16
Avoid What must not appear Extra fingers, invented logos, text on the product
Prompt pack for AI video and UGC ad creators
AI UGC Ad Video Prompts for Sora, Veo and Kling: Product Demo and Hook Scripts
Shot-by-shot prompts and 15-30 second scripts for UGC-style product ads you can generate with AI video tools.
$24 · See what's inside →

Copy-paste prompts

Three prompts from the pack, trimmed to the essentials. Replace every [BRACKETED PART] with your own details before you generate.

When to use: the first three seconds, when you need a hook that works with no dialogue at all.

Vertical 9:16 video, 4 seconds.
Subject: [PRODUCT, WITH COLOR, MATERIAL AND SIZE].
Action: the product drops from above into frame, lands on [SURFACE], settles with a small bounce, then the camera slowly pushes in.
Setting: [SETTING, E.G. BRIGHT MODERN KITCHEN, BLURRED BACKGROUND].
Camera: macro close-up, 50mm look, locked off for the drop, then slow push-in, shallow depth of field.
Lighting: [LIGHTING, E.G. SOFT MORNING WINDOW LIGHT FROM THE LEFT].
Style: crisp, realistic product commercial, natural colors.
Sound: a light landing sound, no music.
Leave the top third of the frame clear for on-screen text.
Avoid: warped or melting product shape, extra products, invented text or logos, hands, people.

If a bracketed detail is missing, ask ChatGPT, Claude or Gemini, or ask your client, before you generate. Do not let the tool fill the gap for you.

When to use: the demo, when the buyer needs to see the product do one thing well.

Vertical 9:16 video, 6 seconds.
Subject: hands using [PRODUCT, WITH COLOR AND MATERIAL].
Action: one clear step only: [THE STEP, E.G. "PRESSES THE TOP BUTTON TWICE AND THE LID POPS OPEN"]. Movements are slow and deliberate.
Setting: [SURFACE AND BACKGROUND].
Camera: close-up from a 45-degree angle above, static or very slow push-in, 50mm look, product in sharp focus.
Lighting: [E.G. SOFT, EVEN, BRIGHT DAYLIGHT, NO HARSH SHADOWS].
Style: realistic tutorial-style UGC, natural colors.
Avoid: complex finger movements, extra fingers, product parts that do not exist, steps that skip ahead, text on screen.

If the tool struggles with the step, film your real hands for this shot and use AI for the hook and B-roll.

When to use: the explainer, when you need a face and a voice without booking a creator.

Vertical 9:16 video, 7 seconds.
Subject: AI-generated presenter, [AGE RANGE, LOOK, CLOTHING], sitting [ON A COUCH / LEANING ON A KITCHEN COUNTER], holding [PRODUCT].
Action: talks to the camera like explaining to a friend, small natural hand gestures, glances at the product once.
Spoken line: "[LINE, UNDER 20 WORDS, WRITTEN AS THE BRAND OR A CREATOR, NOT AS A CUSTOMER]"
Setting: [ROOM DETAILS].
Camera: medium close-up, phone on a tripod at eye level, static, 26mm look.
Lighting: [E.G. SOFT DAYLIGHT FROM A WINDOW TO THE SIDE].
Style: realistic creator video, natural pauses, relaxed and believable.
Avoid: lip-sync drift, stiff posture, reading-from-a-script look, product changing shape, claims of personal results.

An AI-generated presenter should speak as the brand or a creator, never as a made-up customer. Check your platform's labeling rules before you publish.

Worked example

A weak prompt asks the model to do four jobs at once and gives it nothing to hold on to:

Make a cool ad for my reusable coffee cup. Show someone using it at work. Make it look premium and show how it keeps coffee hot. 15 seconds.

The same idea, written as a shot list:

Vertical 9:16 video, 5 seconds.
Subject: hands holding a matte sage-green insulated cup with a bamboo lid.
Action: the person twists the lid off, pours coffee from a small pot, replaces the lid, and sets the cup next to a laptop.
Setting: home office desk, blurred bookshelf behind.
Camera: close-up from 45 degrees above, slow push-in, 50mm look.
Lighting: soft daylight from a window on the left.
Style: realistic UGC, natural colors, no cinematic grading.
Avoid: steam that looks fake, extra fingers, a brand logo on the cup, text on screen.

The second prompt names the subject, one action, the setting, the camera, the lighting, the style, the duration and what to avoid. It also drops the unsupported claim about heat and the vague word "premium." The model now has one job per clip instead of four.

Common mistakes

  • Writing a one-line prompt and hoping for the best. Fix: use the eight-part structure every time, even for a three-second hook.
  • Letting the tool invent features your product does not have. Fix: put the inventions in the "Avoid" line and generate several versions so you can discard the drift.
  • Putting words in the mouth of a fake customer. Fix: have the presenter speak as the brand or a creator, and follow your platform's AI-content labeling rules.
  • Generating one version and shipping it. Fix: generate two to four per clip and keep the ones where the product looks accurate.
  • Running the claims check after you generate. Fix: run the claims and disclosure check on the script first, then generate.
  • Expecting a text-only prompt to match your product exactly. Fix: use the image-to-video prompt with your own product photo, and film real close-ups where the details matter.

Frequently asked questions

Which AI video tools do these prompts work with?

They are written for text-to-video and image-to-video tools such as Sora, Veo, Kling and Runway. The structure (subject, action, setting, camera, lighting, style, duration, avoid) carries across all of them. Tools differ in clip length, audio and lip-sync, so tools without built-in speech need the spoken lines added later as voiceover.

Will Kling keep my product looking like the real one?

Not always. Text-only prompts can change shapes, colors or labels. For accuracy, use the image-to-video prompt with your own product photo, generate several versions, and film your real product for any close shot where the details matter. Keep the versions where the product looks accurate.

Can I use AI-generated presenters in ads?

Many brands do. The presenter should speak as the brand or a creator, not pretend to be a real customer, and you should follow your platform's ad and AI-content labeling rules. The pack includes a claims and disclosure check. That is practical guidance, not legal advice.

Do I need to disclose that a clip is AI-generated?

Follow the rules of the platform you advertise on and the market you sell in; they change, and this guide is not legal advice. As a practical rule, label AI-generated ad content when the platform asks for it, and never present an AI presenter as a verified customer.

What do I get in the pack?

Thirty text prompts in five sections, each with a title, a when-to-use line, the prompt with fill-in placeholders and a filled-in example. You also get a claims and disclosure check. No video files or software are included. You get the text you paste into your own tools.

Related guides

Next step

The pack AI UGC Ad Video Prompts for Sora, Veo and Kling: Product Demo and Hook Scripts has 30 prompts across planning, hooks, demos, talking shots and B-roll, plus the claims and disclosure check. Start with the Product-to-Angle Brief, then generate one hook and one demo for a single product this week.

Get the full pack

More guides

Free prompts.

Two from every pack, as one PDF. Then one email when a new pack is out.

We send one confirmation link first.