AI Model Reviewer

GPT-6 Astra in OpenCode 1.18.32

skills on · OpenCode 1.18.32 + default skills kitKit: default (46 web/Three.js/QA + 33 video) + finish-the-job addendum · 80 skills installed; skill use not recorded

← 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.

OKPower grid daymax effortNeutral harness

Wall time 28m 28sCost $6.12 reported

Review

No verdict yet.

The full review

OK. Finished on its own after 28.5 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, 3,199 characters. Every result for this prompt

Build an interactive electricity grid dashboard for one day on a regional power grid from the exact dataset below (CSV: hourly average generation by source in megawatts, battery output where negative values mean the batteries are charging, total demand, and the wholesale price in US dollars per megawatt-hour). Embed the data in the page verbatim and compute every number from it; do not invent or alter any value.

hour,solar_mw,wind_mw,hydro_mw,gas_mw,coal_mw,battery_mw,demand_mw,price_usd_mwh
00:00,0,908,477,2389,1134,0,4908,66.39
01:00,0,1029,489,1949,1187,0,4654,56.87
02:00,0,1314,469,1408,1216,0,4407,45.61
03:00,0,1416,389,1348,1155,0,4308,48.33
04:00,0,1444,389,1400,1217,0,4450,50.12
05:00,0,1420,472,1781,1150,0,4823,53.61
06:00,0,1364,425,2478,1100,0,5367,70.84
07:00,574,1275,457,2517,1149,0,5972,72.42
08:00,1115,1114,449,2543,1186,0,6407,69.67
09:00,1591,1034,421,2210,1181,0,6437,65.77
10:00,1975,737,428,1798,1154,0,6092,57.38
11:00,2244,644,520,1345,1161,-256,5658,43.57
12:00,2383,422,503,1220,1150,-309,5369,-5.25
13:00,2383,407,517,1047,1182,-287,5249,-6.24
14:00,2244,313,465,1324,1239,-345,5240,48.61
15:00,1975,459,430,1334,1158,0,5356,44.95
16:00,1591,448,380,2092,1205,0,5716,60.08
17:00,1115,697,510,2375,1220,431,6348,65.74
18:00,574,809,515,3503,1199,312,6912,133.18
19:00,0,1027,408,3908,1164,405,6912,139.23
20:00,0,1161,491,3178,1124,394,6348,82.08
21:00,0,1326,487,2768,1135,0,5716,74.8
22:00,0,1226,477,2481,1171,0,5355,71.86
23:00,0,1411,501,2088,1231,0,5231,64.7

The dashboard needs: a stacked area chart of the generation mix over the 24 hours with demand drawn as a line on top and battery charging shown below the zero axis; KPI cards for total energy in MWh, renewable share of generation, peak demand and its hour, and the daily average and minimum price; a price line chart with negative-price hours shaded and labelled; a donut of the daily energy mix by source; a live 'grid now' panel driven by an hour scrubber that shows each source as an animated flow into the demand bar; source toggles that update every chart; hover or tap tooltips with exact values; and a short auto-written insights panel computed from the data (for example when solar exceeded half of demand, when batteries charged and discharged). Style it like a control-room display: dark, precise, glowing. So the page is alive without input, on load the charts draw in and the hour scrubber plays through the day on a loop with the flow panel animating, tooltips following the playhead and the insights highlighting as their hours pass, forever, until the user touches a control.

Save it as index.html in the current directory: one self-contained file with inline CSS and JavaScript only, no libraries or frameworks, and no external requests of any kind (no CDNs, web fonts, images or audio files). It must work when opened directly from disk, fit any window from a 390 px wide phone to a desktop, and run without console errors. You may serve it locally and open it in headless chromium (/usr/bin/chromium, software WebGL, no GPU) 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.

How it was made

Wall time 28m 28sFirst artifact at 9m 43s7 saved versions

Versions saved during the run

7 intermediate outputs, in the order the model wrote them.

Version 7 of 7
  1. Version 1 at 9m 50s thumbnailv1 9:50
  2. Version 2 at 11m 06s thumbnailv2 11:06
  3. Version 3 at 13m 48s thumbnailv3 13:48
  4. Version 4 at 14m 29s thumbnailv4 14:29
  5. Version 5 at 22m 19s thumbnailv5 22:19
  6. Version 6 at 23m 40s thumbnailv6 23:40
  7. Version 7 at 25m 22s thumbnailv7 25:22

No transcript was delivered with this result.

Cost breakdown

Headline cost $6.12, as reported by OpenCode (its own bill for the run). The list-price estimate from the token counts is $5.87.

Tokens by typeTokens$ per 1MCost
Input (uncached)120$10.00$0.0012
Cache write99,242$10.00$0.9924
Cache read2,629,927$1.00$2.6299
Output44,934$50.00$2.2467
of which reasoning11,408billed as output
Total computed from tokens
$5.8702
Total reported by the harness
$6.1184
Difference
Harness-reported $6.1184 vs list-price estimate $5.8702 (-4.1%); the harness figure is the headline.

OpenCode 1.18.32 reported its own total; token counts x list prices shown beside it. Cache writes priced at the input rate (the price table has no separate cache-write price). The harness reports reasoning tokens separately from output; here output includes them (33,526 visible + 11,408 reasoning), and both bill at the output rate.

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

Run notes

one prompt, agentic harness (self-review allowed)

  • frozen prompt (written): written 2026-10-05 for the MR category expansion (owner request: 500 finished runs in every category); category data_dashboards; zero results
  • plain OpenCode config copy (providers only; no ECC instructions/agents)
  • default skills on: 80 skills in this run's xdg/config/opencode/skills
  • opencode export written to a file (not piped)
  • Mode: interactive TUI in tmux (160x48), inside bubblewrap sandbox
  • Isolation audit: outside paths seen
  • Renderer: inline SVG (first <svg> at index.html line 35) in one self-contained HTML file; no libraries, no CDN, no external assets. No WebGPU or WebGL needed.
  • Card video: the run's own video.mp4 (H.264 yuv420p, unchanged). Poster: the run's own render.png.
  • Same frozen prompt as the other rows on this test; configurations differ and all rows are unscored, so this is not a head-to-head result.