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. Tap inside for sound.

Recorded from the live demo. Preview video recorded by this site from the live demo. The moving picture above is that preview; the button runs the real page.

OKVenetian canal watercolour, painted live (HTML)n/a effortRaw baseline

Wall time 7m 27sCost not recorded: the run's token counts were never reported, so there is no estimate

Review

No verdict yet.

The full review

OK. finish=completed

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

raw API call, 1,980 characters. Every result for this prompt

Create a single self-contained HTML file (inline CSS and JavaScript only; no external libraries, fonts, images or network requests) that paints a detailed watercolour painting live, step by step, the way a real artist works.

Scene: a Venetian canal at golden hour. A gondola with a gondolier in the foreground, an arched stone bridge in the middle distance, tall old buildings with balconies, shutters, flower boxes and hanging laundry on both sides, a church dome and bell tower against a warm sky, soft reflections rippling in the water.

The painting builds itself on a canvas and/or SVG over about 25 seconds, in four visible stages:
1. Pencil sketch (about 0-6 s): light grey graphite construction lines and contours drawn stroke by stroke, slightly wobbly and hand-made.
2. Ink lines (about 6-11 s): confident dark brown/sepia pen lines trace over the key contours, stroke by stroke.
3. Layered washes (about 11-21 s): transparent watercolour washes bloom in, from light to dark: sky first, then buildings, water, shadows. Soft bleeding edges, pigment pooling at the edges of each wash, granulation, visible paper texture showing through.
4. Details (about 21-25 s): final darker accents, window details, highlights on the water, a few loose splatters.
Then hold the finished painting on screen (do not loop or restart).

Requirements:
- Start drawing immediately at page load; the first strokes must be visible within the first half second.
- Portrait 4:5 composition on warm off-white textured paper, scaled to fit the browser window and centred; the page background around it is a neutral colour.
- Drive all animation with requestAnimationFrame and performance.now (no setTimeout-only timing, no user interaction needed).
- No text, titles, labels, captions, buttons or UI of any kind on the page. Only the painting.
- Make it as beautiful and detailed as you can: it should look like a real watercolour, not flat vector shapes.

Reply with only the complete HTML file.

How it was made

Wall time 7m 27s

No transcript was delivered with this result.

Cost breakdown

No cost figure: the run's tokens were never reported, so there is nothing to estimate from.

Tokens by typeTokens$ per 1MCost
Input (uncached)not recordedn/an/a
Cache writenot recordedn/an/a
Cache readnot recordedn/an/a
Outputnot recordedn/an/a
Total computed from tokens
unknown
Total reported by the harness
none