All Samples
AI Art Directors·3 tools · 9 outputs
Kokeshi Toy Style
A Japanese kokeshi dolls aesthetic applied to AI image generation and web design — no re-explaining, ever.StyleRef encodes the visual or tonal style from the source reference into a structured format — reusable across every prompt, every generation.
This StyleRef is prepended to every prompt shown below — that's what keeps every output consistent.
About This Style
The Kokeshi Toy Style demonstrates how a single StyleRef can lock a highly specific aesthetic across completely different AI tools — ChatGPT's image generation and Gemini's multimodal output — without re-explaining anything.The source reference is a set of Momiji message dolls: round geometric forms, a restricted palette of muted earth tones and soft pinks, flat studio lighting, and clean negative space. Every one of those characteristics is encoded into a reusable StyleRef block, covering shape language, color palette (with exact hex values and percentage weights), surface treatment, lighting, and spatial hierarchy.The result: you can prompt "Design a rabbit toy" or "A wooden dinosaur toy" into any compatible AI tool and get outputs that feel like they belong to the same product family. Same visual DNA, different subjects.This style is particularly useful for AI Art Directors, product designers, and anyone producing multiple AI-generated assets that need to feel cohesive — toy packaging, character design sheets, children's app illustration, and themed merchandise all benefit from this kind of constrained aesthetic.
03 — AI Outputs
Same StyleRef. Different Prompts. Consistent Results.
Generated with
ChatGPT
— Image GenerationGenerated with
Gemini
— Image GenerationGenerated with
Gemini
— Web DesignBuild a StyleRef for your style
Define it once from your references. Paste it into your AI tool. Get consistent results.Create Your StyleRef — FreeNo account needed© 2026 StyleRef. All rights reserved.









