AI Model Reviewer

GPT-6 Astra in a raw API call

raw API baseline

← Open in the gallerySame prompt, all models

Zoom, inline player and more ways to view
Poster frame of GPT-6 Astra'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).

OKRube Goldberg breakfast machine on a kitchen countertop (Three.js)n/a effortRaw baseline

Wall time 5m 39sCost $0.65 estimated

Review

No verdict yet.

The full review

OK. finish=completed; ran in the capture with no console errors

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, 1,049 characters. Every result for this prompt

Build a Three.js scene: A pastel kitchen countertop holds a chain-reaction contraption: a rolling marble tips dominoes, which trigger a seesaw, release a toy car down a ramp, pull a string that drops toast into a toaster, and finally flip a lever that pours a cup of cereal. Click to start the sequence and press R to reset it, and each stage must visibly cause the next. A cinematic camera follows the active stage along a smooth spline path. Everything is built in code from Three.js geometry, materials and shaders; no external models, textures or fonts. The scene must visibly animate from the moment it loads and run in a browser with software WebGL (no GPU). Deliver ONE self-contained HTML file; load Three.js only through an import map from https://cdn.jsdelivr.net/npm/three@0.170.0/build/three.module.js (addons from https://cdn.jsdelivr.net/npm/three@0.170.0/examples/jsm/), with no other external requests. Output ONLY the complete HTML file, starting with <!DOCTYPE html> and ending with </html>, with no explanation or markdown fences.

How it was made

Wall time 5m 39s

No transcript was delivered with this result.

Cost breakdown

Headline cost $0.65, estimated from the token counts at list price.

Tokens by typeTokens$ per 1MCost
Input (uncached)235$10.00$0.0024
Output13,003$50.00$0.6502
of which reasoning2,662billed as output
Total computed from tokens
$0.6525
Total reported by the harness
none

Computed from provider-reported token counts x list prices.

Prices: Project price table (vendor list prices), 2026-09-22.

Run notes

one prompt, raw API (single request, no tools)

  • Renderer: WebGL via three.js WebGLRenderer (three@0.170.0 loaded from the cdn.jsdelivr.net CDN through an import map plus three.js addons from the same CDN); text/HUD overlay in DOM+CSS. No local files.
  • Card video: the row's own capture, re-encoded as H.264. Poster: a frame of that capture.
  • Same frozen prompt as the other rows on this test; all rows are unscored, so this is not a head-to-head result.