AI Model Reviewer

GPT-6 Sol in Codex CLI 0.156.1

skills off

← Open in the gallerySame prompt, all models

Zoom, inline player and more ways to view
Poster frame of GPT-6 Sol's page
Open the original page

The model's own page, run in a sandboxed frame only when you ask (a page that needs its own origin opens in a new tab on a separate demo site). Phones get a full-screen player capped at 1.5x pixel density (2x on larger screens) so it stays smooth; the original page runs uncapped. It starts in its auto-tour (T switches to play mode on a keyboard).

OKRocket car arenamax effortNeutral harness

Wall time 7m 10sCost $1.03 estimated

Review

No verdict yet.

The full review

OK. Finished on its own after 7.2 min.

Scores

AI judge

Not scored

No rubric for this prompt

Visitor votes and reviewer grades are not available yet.

AI judge
AI judge: the test runner (model claude-opus-5-5), scoring by visual review of the 1600 px render against the pre-registered rubric. No AI-judge score exists for this result: there is no rubric for this prompt, so no score is shown and none is made up.
Visitors
Public voting with a blind reveal is coming: you score the output first, then see the other scores.
Reviewer
The reviewer's own grade appears here once it is added.

Scores compare results for the same prompt only. AI-judge boards rank runs within one prompt, equal totals sharing a rank; scores from different prompts are never averaged.

Downloads

Files actually published for this run. Original output, site-made previews and post-run conversions are labelled separately.

The exact prompt

as run, 1,087 characters. Every result for this prompt

Build a small Three.js "rocket car soccer" arena game in the browser: an enclosed stadium with ramped walls, a big physics ball, two goals, and a player car that can drive, jump and boost (boost leaves a particle trail). Simple arcade physics written by hand (no external physics engine download needed, but a CDN physics lib is allowed). Score UI, a 2-minute match timer, a goal replay flash. Controls: WASD/arrows, Space jump, Shift boost, with touch buttons on mobile. IDLE AUTO-PLAY: with no input, an AI drives the car toward the ball, scores, and the chase camera follows, so a headless capture shows gameplay. Must render on WebGLRenderer. Save the main page as index.html in the current directory (plus any local asset files you create), with no build step. You may serve it locally and open it in headless chromium (/usr/bin/chromium, software WebGL, no GPU, no WebGPU) to screenshot and inspect your own output, and iterate on your own before finishing. Close any browser you open as soon as you've taken your screenshot. Finish with a short README.md describing the controls.

How it was made

Wall time 7m 10sFirst artifact at 2m 02s1 saved version

Versions saved during the run

1 intermediate output, in the order the model wrote them.

Version 1 of 1
  1. Version 1 at 2m 04s thumbnailv1 2:04

No transcript was delivered with this result.

Cost breakdown

Headline cost $1.03, estimated from the token counts at list price.

Tokens by typeTokens$ per 1MCost
Input (uncached)99$2.00$0.0002
Cache write74,700$2.00$0.1494
Cache read1,696,564$0.20$0.3393
Output53,689$10.00$0.5369
of which reasoning26,240billed as output
Total computed from tokens
$1.0258
Total reported by the harness
none

Computed from harness-reported token counts x list prices. Cache writes priced at the input rate (the price table has no separate cache-write price). No request above 272,000 input tokens (33 checked): standard rates.

Prices: Project price table, 2026-09-25.

Run notes

one prompt, agentic harness (self-review allowed)

  • frozen prompt (written): X-idea prompt queue (frozen in prompts_queue), MR wave 8 Three.js/WebGL expansion; category threejs; idle auto-play mode in prompt; WebGL-renderable; engine threejs harness suffix appended (the source's own 'Deliver site/index.html' sentence dropped as the suffix carries the save location); zero results
  • plain CODEX_HOME copy (config.toml only, run work/ trusted; no AGENTS.md/skills/state)
  • Mode: interactive TUI in tmux (160x48), inside bubblewrap sandbox
  • Isolation audit: outside paths seen
  • Renderer: Three.js r160 WebGLRenderer (site/game.js line 13). Three.js is vendored as a local ES module (site/three.module.js, imported at game.js line 1; game.js is a module script at site/index.html line 151), so the demo needs no CDN. No WebGPU needed.
  • No card video (the http re-capture has only 5 frames over 8 s on a GPU-less box): the card uses the poster, re-captured over a local http server, because the run's own render.png came from a file:// capture that cannot load the page's local ES-module Three.js and shows an empty viewport.
  • Same frozen prompt as the Claude Code + Opus 5.5 row (threejs_rocket_car_arena__claude-code__opus-5-5-high__attempt1). The two configurations differ (Codex at effort max, Claude Code at effort high) and both rows are unscored, so the pair is not a head-to-head result.