How to make AI ASMR videos for ads: a creator workflow

On this page

What is an AI ASMR ad?
Tools creators use for it
How do you make an AI ASMR ad?
Prompt template for an AI ASMR clip
Common problems and fixes

What is an AI ASMR ad?

An AI ASMR ad is a slow, close-up product video in which texture and sound do the selling: a lid twisting off, a cream pressed out, a drink poured over ice. Every shot and every sound is generated by AI, and there is usually no voiceover.

The format borrows from ASMR videos people watch to relax, so the pace is calm and the camera stays close. In an ad, the product is at the centre of the process: being made, unboxed, poured or applied. We do not have a public examples page for ASMR yet; the nearest format is AI product demo ads, which come from posts on X and were not made on or for Anglr.

Tools creators use for it

The table counts how many of the reviewed tutorials named each tool for ASMR clips. A tutorial that used two tools counts once for each.

Based on 8 public tutorials posted between 8 May 2026 and 30 Sep 2026, reviewed by Anglr.

Tool	What it does in this workflow	Tutorials that used it
Seedance	Animates the storyboard or a single long prompt into the ASMR clip, with sound	7
GPT Image	Draws the storyboard image that plans every shot	4
Omni	Named next to Seedance as an alternative video model	2

Single tutorials used Kling for a cooking scene and Nano Banana as an alternative image model, and three ran their models through hosting apps (OpenArt, Pollo AI and Yapper). The first two storyboard posts in our sample (@Ciri_ai and @Shorelyn_) open with near-identical storyboard prompts a day apart, so they may share one source.

How do you make an AI ASMR ad?

You make it in six steps: choose the process, plan it on one storyboard image, animate slow close-ups, write the sound, add a person only if needed, then cut. The order is our synthesis of the tutorials.

1. Choose a process with texture

Pick a moment where something changes in a satisfying way, and put the product at its centre. The tutorials agree on the type of subject even when the products differ. @Zyrellix shows a cosmetic being made and packed step by step. @Hope_Ai01 films a soup being cooked. @AIwithSynthia and @AIwithkhan use food and drink. @Ciri_ai and @Shorelyn_ use an unboxing.

Write down the sequence before you prompt: closed, opened, revealed, used, finished. Each stage should have one texture and one sound you can name.

2. Plan it on one storyboard image

Four of the tutorials plan the clip as a storyboard image first and only then animate it. @Ciri_ai and @Shorelyn_ ask GPT Image 2 for a clean storyboard infographic of a product unboxing on a white background. @AIwithSynthia makes a landscape storyboard board for a food commercial. @AIwithkhan makes the storyboard with a character reference attached.

The board does two jobs. It fixes the order of shots, and it gives the video model one reference that already shows the product, the setting and the light. Check it panel by panel before spending anything on video: the product should look the same in every frame and the steps should follow each other.

3. Animate slow, close shots

Animate in short, close shots with small movements. The video model gets the board (or the first frame) as its reference and a prompt that describes each shot in order. Seven of the eight tutorials used Seedance for this step.

BytePlus's Seedance prompt guide gives two rules that suit ASMR well: prefer slow, small, continuous movements, and use one camera movement per shot, because combining moves makes the image less stable. Close framing, shallow focus and soft light do the rest; @Hope_Ai01 asks for warm golden light and shallow depth of field.

Three authors skip the board. @Zyrellix, @Hope_Ai01 and @mehvishs25 describe the whole clip in one long prompt, shot by shot. That is quicker for a single clip; the board is easier to reuse across several products.

4. Write every sound into the prompt

Name each sound the way you would name a shot. ASMR lives or dies on audio, and video models that generate sound only make what you ask for. The tutorials we reviewed say little about sound in the post text itself, so this step leans on the vendors' guides.

Model	How to write sounds	Vendor source
Seedance 2.0	Sound effects in angle brackets, music in parentheses, speech in curly braces	BytePlus guide
Veo 3.1	Sound effects after SFX:, the background after Ambient noise:	Google Cloud guide

Be specific: "a soft crackle as the foil seal peels back" gives the model more to work with than "peeling sound". For a pure ASMR ad, leave music out or keep it very low, so the product sounds carry the clip.

5. Add a person only if the format needs one

Most ASMR ads show hands and product, not faces. When a person does appear, lock them. @AIwithkhan opens the prompt by listing every feature to keep from the character reference (face, hair, eye colour, skin tone, makeup). @mehvishs25 mixes a UGC-style creator with ASMR product sounds. @Zyrellix makes the person a 3D cartoon in a different variant. See the AI 3D animated ads guide for that look.

6. Cut with sound in mind

Edit for rhythm: let each sound finish before the next shot starts, and avoid cutting in the middle of a pour or a peel. Add the product name and the offer as on-screen text near the end, since the clip usually has no voiceover. Check that every frame shows the same product, with no extra or missing parts, before it goes out. The AI product videos guide covers product consistency in more depth.

Prompt template for an AI ASMR clip

These templates are ours. The storyboard follows Google's Nano Banana guide (state the purpose, be specific, put any text to render in quotation marks); we have not reviewed a GPT Image prompt guide, so we use the same order there. The motion prompt follows BytePlus's Seedance order, with shots labelled in sequence, sound in angle brackets and constraints at the end. Duration and aspect ratio go in the settings. Build your own in the AI ad prompt builder, or see Seedance prompts.

Storyboard (image model, with the product photo attached):

A clean storyboard sheet for a short ASMR product video, six panels in two rows on a white background, each panel numbered. The product in every panel is exactly the one in the attached photo: same shape, colour and label. Panel 1: [product] closed on [surface], soft side light. Panel 2: hands [first action: unscrew, peel, tear]. Panel 3: macro of [texture revealed]. Panel 4: [product] being [poured, pressed, applied]. Panel 5: [result close-up]. Panel 6: [product] upright, label facing the camera. Macro lens look, shallow depth of field, warm soft light throughout.

Motion (Seedance, image-to-video with the storyboard as reference):

Image 1 is the storyboard. Image 2 is the product. Follow the panels of Image 1 in order.
Shot 1: macro, static camera. [Product] from Image 2 rests on [surface]. <a quiet room tone>
Shot 2: slow push in. Hands slowly [first action]. <[precise sound, e.g. a soft crackle as the seal lifts]>
Shot 3: static macro. [Texture] [slow small movement, e.g. folds and settles]. <[precise sound]>
Shot 4: slow pan right. [Product] is [poured, pressed, applied] in one continuous motion. <[precise sound]>
Shot 5: static. [Product] upright, label facing the camera.
Keep the product shape and label the same as Image 2. Keep it subtitle-free. Do not generate a logo or a watermark.

Common problems and fixes

Problem	Fix	Source
Shots come out in the wrong order	Plan the clip on one storyboard image and give it to the video model	@Shorelyn_, @Ciri_ai
The person's face changes between shots	List every feature to keep from the character reference at the top of the prompt	@AIwithkhan
The image shakes or warps	Use one camera movement per shot	BytePlus Seedance guide
Movement looks jerky	Ask for slow, small, continuous movements	BytePlus Seedance guide
The sound is generic	Name each sound effect precisely, in the model's sound syntax	Google Cloud Veo guide

ASMR has fewer published methods than UGC or commercials, and none of the posts we reviewed reports results for an ASMR ad, so there is no "What creators report" section here. If you already make ASMR or product ads that perform, the creators page explains how Anglr pays for them.

Who this is for

Independent AI creators and editors. You make the ads yourself. You know your work performs, so you'd rather be paid on what it does than on a flat fee.
AI agencies and studios. You run a team of editors. You write the brief, your team makes the versions, and you're paid on performance because you know your team's ads scale.

Not for you if

You only want a flat fee.
You're new to editing and not yet confident your ads will perform.
You want every shot spelled out for you.
You make one style and won't try others.

FAQ

What is an AI ASMR video?

It is a short video, made with AI image and video models, that focuses on close-up textures and quiet, crisp sounds: pouring, cutting, peeling, tapping, unwrapping. In ads, the product is the thing being opened, poured or made.

How do you make AI ASMR videos?

Pick a process with strong textures, plan it as a storyboard image, then animate it in a video model with slow, small movements and every sound written into the prompt. Generate several takes and keep the ones where motion and sound line up.

Can AI generate ASMR sounds?

Yes. Video models such as Seedance and Veo generate audio from the prompt. Seedance's guide puts sound effects in angle brackets; Google's Veo guide prefixes them with SFX:, so name each sound as precisely as you would describe a shot.

Which AI is best for ASMR videos?

In the tutorials we reviewed, Seedance was used for almost every ASMR clip, usually with GPT Image making the storyboard first. Kling was used once, for a cooking scene.

Do ASMR ads work for every product?

They suit products with a tactile moment: food and drink, cosmetics, packaging worth opening, anything poured, pressed or peeled. A product with nothing to touch or hear has little for the format to show.

Sources

Checked: 12 Oct 2026