AI Model Reviewer

GLM-5.3 in OpenCode 1.18.32

skills off

← Open in the gallerySame prompt, all models

Zoom, inline player and more ways to view
GLM-5.3's SVG for Lighthouse on a headland at night

OKLighthouse on a headland at nightmax effortNeutral harness

Wall time 13m 48sCost $0.75 reported5.5/10 · Draft score — AI judge

Review

The idea is there: a dark tower with a glowing lantern and a beam on a rocky point, a moon, and dotted reflections. But the whole image is covered by dense wavy white squiggles (the rain pattern is heavily distorted), which swamp the scene.

The full review

The idea is there: a dark tower with a glowing lantern and a beam on a rocky point, a moon, and dotted reflections. But the whole image is covered by dense wavy white squiggles (the rain pattern is heavily distorted), which swamp the scene. The sea is calm, with no breaking waves, and the headland is barely readable.

Verdict by the AI judge (provisional).

OK. Finished on its own after 13.8 min and produced lighthouse.svg.

Bugs

  • rain pattern renders as dense wavy squiggles over everything
  • calm sea
  • headland barely readable

Scores

AI judge

5.5/10

Draft score — AI judge

Visitor votes and reviewer grades are not available yet.

CriterionMarks
lighthouse1.5
stormy sea + moonlight1
beam through rain/mist1.5
atmosphere1
craft0.5
Total (sum of the marks)5.5 / 10
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. A draft score until the reviewer checks it. Rubric: Wave-1 SVG scoring rubric (AI Model Reviewer).
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

agentic harness (save the file; self-render allowed), 378 characters. Every result for this prompt

Create an SVG of a lighthouse on a rocky headland at night: stormy sea, the lighthouse beam cutting through rain and mist, moonlight on the waves, as atmospheric and painterly as you can. Save it as lighthouse.svg in the current directory. You may render and inspect your own output (headless chromium is available at /usr/bin/chromium) and iterate on your own before finishing.

How it was made

Wall time 13m 48sFirst artifact at 8m 06s6 saved versions

Versions saved during the run

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

Version 6 of 6
  1. v1 7:54
  2. v2 10:50
  3. v3 11:32
  4. v4 11:43
  5. v5 12:35
  6. v6 13:56

Step by step

The prompt, every agent turn and tool call (tool name and a one-line summary; inputs and outputs are not shown), then the final message. The full transcript is 3.1 MB; it is published as a scrubbed, compressed download rather than shown on the page.

Show all 79 steps

Loading the timeline…

Download the full transcript OpenCode session export, 83 KB, gzip. Scrubbed: anything that looks like a key, token, e-mail address or phone number is replaced, and account details and the test machine's file paths are removed.

Cost breakdown

Headline cost $0.75, as reported by OpenCode (its own bill for the run). The list-price estimate from the token counts is $0.75.

Tokens by typeTokens$ per 1MCost
Input (uncached)79,349$1.40$0.1111
Cache write0$0.00$0
Cache read1,444,088$0.26$0.3755
Output25,335$4.40$0.1115
reasoning35,254$4.40$0.1551
Total computed from tokens
$0.7531
Total reported by the harness
$0.7531
Difference
computed 0.5980 vs harness 0.7531

Tokens from OpenCode per-message usage; harness total = sum of OpenCode per-message cost. OpenCode reports reasoning separately; both bill at the output rate (output line here = output only, see reasoning).

Prices: Fireworks GLM-5.3 serverless price (STATE.md 22:42Z; same as the OpenCode/models.dev catalog entry fireworks-ai/glm-5p3), 2026-09-23.

Run notes

one prompt, neutral harness (OpenCode), self-review allowedwave 1

  • Mode: interactive TUI in tmux (160x48), inside bubblewrap sandbox
  • Isolation: bwrap: private /tmp and /run, private PID ns; only this run dir, harness binaries (ro), opencode config, /usr, /etc bound; HOME/TMPDIR/XDG data private; the project files and other runs not mounted