Character Lab

An experiment in AI-generated character consistency and browser-based image manipulation. Six visual styles, real-time proportion controls, lighting and export, without regenerating the image on every interaction.

3D Toy·1/6

Style

Proportions

Head size100%

SmallerBigger

Body0

LeanSturdy

Height100%

ShorterTaller

Lighting

Background

How it works

Character Lab explores AI-generated character consistency across six visual styles, then uses browser-side transformations for real-time editing without regenerating the image.

Generated once

The six styles are pre-generated images of the same character. Switching between them never calls an AI model.

Edited in the browser

Head, body and height are canvas transformations: the figure is split into head, torso and legs bands and each band is scaled smoothly, so the silhouette stays intact.

Processed in your browser

Character edits and PNG exports are processed locally in your browser. Your edits are never uploaded and no credit is spent.

GPT Image
Claude
Next.js
React
GSAP
Canvas 2D