On this page
What is an AI claymation ad?
An AI claymation ad is a short ad generated to look like stop-motion clay animation: modelled characters and sets, soft fingerprinted surfaces, warm light and slightly stepped movement. Nothing is sculpted or filmed; image and video models produce the whole thing.
The handmade look is the point. It reads as craft rather than as AI, and it stands apart from selfie UGC in the same feed. Public examples of clay and other craft looks are on AI claymation and craft ads; they come from posts on X and were not made on or for Anglr.
Claymation is a close cousin of 3D animation. The AI 3D animated ads guide covers the shared workflow (look, script, character sheet, frames, motion) in more depth; this guide covers what is different about clay.
Tools creators use for it
The table counts how many of the reviewed tutorials named each tool for claymation-style ads. A tutorial that used two tools counts once for each.
Based on 6 public tutorials posted between 13 Mar 2026 and 16 Sep 2026, reviewed by Anglr.
| Tool | What it does in this workflow | Tutorials that used it |
|---|---|---|
| Seedance | Animates the clay scenes, as one multi-scene clip or shot by shot | 3 |
| Claude or ChatGPT | Writes the script and voiceover, or runs a skill that builds the clips | 2 |
| GPT Image | Makes the stills and reference images | 2 |
| Higgsfield | Runs the image and video models in one place | 2 |
Single tutorials also used Runway through its MCP connection, Kling, Veo, Nano Banana and MaxFusion. This is a small sample: claymation has few published methods, and two of the six posts name tools without giving the steps, so most steps below are attributed to one author by name.
How do you make an AI claymation ad?
You make it in five steps: choose what the clay look is for, write the script and voiceover, fix the clay description, generate the scenes, then cut to the narration. The order is our synthesis; where only one author describes a step, we say so.
1. Decide what job the clay look does
Each author uses claymation for a different job, so pick one before you write.
- Explainer. Ori Zilbershtein uses it to explain how a product works, why earlier solutions failed, and to build up a problem for buyers who know they have one but not the fix.
- Narrated story. @Aducate builds clay stories around a documentary-style narration.
- Product character. @yining_shi gives the product a small cast (animal characters) and a pun line.
- Process. @Soaima_Ai shows an object being built step by step out of clay.
- Real product, clay world. Peter Quadrel keeps the real product and generates only the world around it.
The explainer and the narrated story need a voiceover; the process and character versions can run on sound effects and a single line.
2. Write the script and voiceover first
Three authors put the words before any image. @Aducate has ChatGPT or Claude write a 60-second documentary-style script and voiceover as the first step. Ori Zilbershtein's Claude skill asks a few questions and then builds the clips from the answers. Muhammad Basit writes the story as timed shots (hook, problem, demo, payoff, brand) before building anything visual.
Claymation is slow to read, so give each scene one idea and let the narration carry the argument. Read the script aloud at the pace of the voice you will use; the number of scenes follows from that. Our UGC script template covers the hook and call-to-action beats.
3. Fix the clay description and the materials
Write one fixed description of the look and reuse it word for word in every prompt. Peter Quadrel locks a style first so new variants are quick. @Soaima_Ai's prompt asks the model to keep the visual style consistent across all scenes. Muhammad Basit builds character, product and environment sheets, with materials and lighting written down, before any frame is generated.
For clay, the description should name the material, not just the style: modelling clay with visible fingerprints and tool marks, slightly uneven edges, matte surfaces, small handmade props, soft warm studio light, a shallow tabletop set. Name the colour palette once and keep it. Describe the movement too: stop-motion reads as small steps rather than smooth motion.
Then make a reference image of each character and the main set in an image model, and attach it to every scene. Google's Nano Banana guide suggests stating the purpose of the image and being specific, which helps here: "a clay character sheet for a stop-motion ad" gets a more usable result than "a clay character".
4. Generate the scenes
The authors split on how. @Soaima_Ai writes one Seedance 2.0 prompt that lays out a short clip as a run of numbered scenes, with GPT Image for the stills. @yining_shi asks Runway for the whole ad in plain language through its MCP connection inside a chat. Muhammad Basit approves the boards first, then makes start and end frames and directs each shot's motion in Seedance.
A single multi-scene prompt is fastest for a short process clip. Shot-by-shot generation from approved frames gives more control, and it suits narrated explainers where each scene has to match a line of the voiceover.
BytePlus's Seedance guide gives two rules worth following: label shots in the order they happen, and use one camera movement per shot. It also advises against per-shot timings in the prompt, because timing support is unstable; set the length in the settings and let the edit do the timing.
5. Cut to the narration and check the claims
Lay the voiceover down first, then cut the scenes to it, and add small sound effects for clay moving, squashing and settling. Check that each character and the product look the same from scene to scene.
Claymation is popular for mechanism explainers in health and supplement ads, and the friendly look does not change the rules. Every claim in the voiceover must be one the brand can legally make, and animation of the body or of a product "working" should not show results the brand cannot support. Anglr's creators page sets out the standards every ad must meet.
Prompt template for an AI claymation scene
This template is ours. It follows BytePlus's Seedance order: subject, action, setting, light, camera and style, with shots labelled in sequence, sound in angle brackets, speech in curly braces 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.
Character sheet (image model):
A character sheet for a stop-motion clay animation ad. [Character: a small round clay figure, colours, one distinctive feature]. Made of matte modelling clay with visible fingerprints and small tool marks, slightly uneven edges. Front, side and three-quarter views on a plain light grey background, soft warm studio light. Keep the shape and colours identical in every view.
Scenes (Seedance, with the character sheet and a product photo attached):
Image 1 is the character. Image 2 is the product.
Style for every shot: stop-motion claymation, matte modelling clay with fingerprints, handmade tabletop set, soft warm light, slightly stepped movement.
Shot 1: static wide shot. The character from Image 1 sits on a [clay set] looking [problem emotion through posture: slumped shoulders]. <a soft clay squish>
Shot 2: slow push in. A clay version of [the problem] appears beside them. <[sound]>
Shot 3: static medium shot. [Product] from Image 2 slides into frame; the character picks it up. They say {[one short line]}.
Shot 4: static wide shot. The set changes to [the result], the character stands tall. <[sound]>
Keep the character the same as Image 1 and the product the same as Image 2. Keep it subtitle-free. Do not generate a logo or a watermark.
Common problems and fixes
Few of the posts state problems, so most of the fixes here come from the vendor guide or are attributed to one author.
| Problem | Fix | Source |
|---|---|---|
| Scenes drift into different styles | Keep one fixed style description in every prompt | @Soaima_Ai, Peter Quadrel |
| The clip comes out before the story is clear | Write the script and voiceover before generating | @Aducate |
| Frames look wrong after video credits are spent | Approve the boards and stills before animating | Muhammad Basit |
| Scene timing ignores the prompt | Drop per-shot timings; label shots in order and time them in the edit | BytePlus Seedance guide |
| The image wobbles | Use one camera movement per shot | BytePlus Seedance guide |
What creators report
These are the authors' own claims, not results we checked.
- Ori Zilbershtein reports that a claymation explainer takes an experienced maker about three hours by hand and about ten minutes with the Claude skill.
- @Aducate reports that claymation storytelling videos are doing very well on the accounts the author works on.
If you already make animated or craft-style 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
Can AI make claymation videos?
Yes. Image and video models can imitate the look of clay stop-motion: soft modelled shapes, fingerprints in the surface and slightly stepped movement. Creators generate scenes in models such as Seedance or Runway and finish them in an editor.
How do you make an AI claymation video?
Write the script and voiceover first, describe the clay look the same way in every prompt, make a reference image of each character or set, then generate the scenes one by one or as a short multi-shot clip. Lay the narration and sound over the cut.
What is the best AI for claymation?
The posts we reviewed used Seedance most, with Runway, Kling and Veo also named. The fixed description of the clay look matters more than the model, because it is what keeps every scene in one style.
What are claymation ads good for?
Explaining how something works, telling a short story, and giving a product a character. Ori Zilbershtein uses them for problem-aware buyers and mechanism explainers; any health or product claim still has to be one the brand can legally make.
How long should an AI claymation ad be?
The posts we reviewed range from a 15-second multi-scene clip to a 60-second narrated story. Plan the story to the length of the voiceover, then cut a shorter version that keeps the problem, the product and the end.
Sources
- @Aducate (13 Mar 2026): claymation storytelling that starts with a 60-second documentary-style script and voiceover from an LLM
- Peter Quadrel (9 May 2026): animation ads that keep the real product inside a generated world, claymation among the styles
- @yining_shi (27 May 2026): plain-language requests to Runway through its MCP for clay product ads with animal characters and a pun
- @Soaima_Ai (16 Aug 2026): a short multi-scene clay stop-motion of an object being built, from GPT Image and Seedance 2.0
- Muhammad Basit (4 Sep 2026): one product turned into several animated worlds, claymation included, after sheets and boards are approved
- Ori Zilbershtein (16 Sep 2026): where claymation explainers fit (mechanisms, failed solutions, problem-aware buyers) and a Claude skill to build them
- BytePlus ModelArk – Dreamina Seedance 2.0 series prompt guide
- Google Cloud – The ultimate Nano Banana prompting guide
Checked: 12 Oct 2026