AI image prompts to convert portrait photos into handmade stop-motion puppet style inspired by the classic Czech animation Pat & Mat. Recreates the charming clumsy clay and felt texture of vintage European children's TV puppets.
2026.06.19
把「Stop-motion Puppet」做成你想要的样子做一个你自己的「Stop-motion Puppet」
用大白话说出来就行——Carat 会为你生成合适的图片和视频。
每天 100 免费积分,现在就开始




Stop-motion puppet AI prompts turn a portrait photo into a handmade puppet in the style of classic European stop-motion animation, most famously the Czech series Pat & Mat. Think a short, chubby 3-heads-tall figure with a simple tan clay face, tiny dot eyes, a big protruding nose, yarn hair, and fabric clothes. With Carat you can generate this look with GPT Image 2 while keeping the outfit colors and hairstyle from the original photo, then refine the result or convert it to other ratios and video afterward.
The charm of stop-motion puppet style lives in its imperfect, handmade texture. If the result looks like a polished fashion doll or a glossy figurine, the whole mood is lost, so the goal is to keep the clay face clumsy and the fabric-and-yarn details slightly rough. Keeping the original number of people, outfit colors, and hair while changing only the art style gives you a puppet that is still instantly recognizable.
With Carat you can create this stop-motion puppet conversion in GPT Image 2 and keep iterating from the result, including ratio changes and video conversion.
A handmade puppet aesthetic inspired by the Czech animation Pat & Mat, with a chubby 3-heads-tall body, clay face, dot eyes, big nose, and yarn hair.
Full-body or half-body portraits work best. A clear, front-facing photo where outfit colors and hairstyle are visible gives the most stable result.
Yes. This is an image-to-image (i2i) edit that references your uploaded photo, so an original portrait is required.
Add negatives such as 'not a cute fashion doll, not a polished toy figure' to keep the clumsy handmade puppet look.
Yes. The prompt keeps the original number of people and their positions and converts the whole group into puppets.
GPT Image 2 in image-to-image (i2i) mode, referencing your uploaded photo to generate the puppet version.
Use Pat & Mat only as a style reference and reinterpret it as an original puppet rather than copying a copyrighted character, which keeps your result safe to use.