Skip to content
VibeCodeBench

Jelly Orange Slice

GPT-6.1 Sol made this.

OpenAIEffort: Medium, level 2 of 5

Made by a Codex agent. It wrote the file in one go and never ran or fixed it.

Jelly Orange Slice by GPT-6.1 Sol: a still of the demo
partly works · auto-checked (Checked by an automated pass. Not hand-played yet.)
Click it to play here.Playing here. Done gives the page back.

Receipt

Made in
Codex
Paid with
ChatGPT subscription
Time
3 min 1 s
Tries
1 try · partly works
Date
Nov 22, 2026
By
a Codex agent
API equivalent (estimate)tokens × list price
about $0.09

Estimate at list API prices. I used my regular paid plan, not the API.

Prompt and answer tokens at list price, thinking included; the coding tool's own instructions are not counted.

Try

  1. Try 1partly works · auto-checked (Checked by an automated pass. Not hand-played yet.)3 min 1 s · 22.0 KBshownCode of try 1
How it was made

One-answer agent run: a Codex agent running GPT-6.1 Sol at medium effort got only the prompt and wrote the file in a single write. It did not open or test it, and nobody edited it afterwards.

What it built, in its own words: Built an interactive 3D citrus jelly on a white plate, with glossy segments, juice cells, peel, and soft-body physics. Supports poking, dragging, squeezing juice drops, firmness adjustment, reset, and a 20-second #demo loop.

Limitations it reported: Requires WebGL and internet access to load three.js; not run or tested, as instructed.

Status not checked by a person yet.

  • Try 1: Automated check: Soft-body drag, Squeeze drops, slider and Reset work, but the jelly reads as a pale cream wedge, not glossy translucent citrus, and the plate is barely visible.

How it was checked

Console errors
1
Links and libraries
all allowed
Hands-free demo
yes
Phone check
390x844
Libraries it loads
cdn.jsdelivr.net/npm/[email protected]/build/three.module.js
See the code

The benchmark

Jelly Orange Slice

A glossy orange slice made of jelly. Poke it, squeeze it, watch it wobble.

I keep the exact prompts private for now.

Every AI on this benchmark