ecumene/rust-sloth | ProgramBench

A 3D software rasterizer... for the terminal!

380

Generated Behavioral Tests

97.9%

Best Score

Claude Opus 5.5 (xhigh)

Results by Model

Hover for details · Click a model to open it · Drag to zoom the x-axis · Shift-drag to pan · The line marks the Pareto frontier (best score per cost)

30 runs

#	Model	Score	Cost	Calls
1	Claude Opus 5.5 (xhigh)	97.9%	$42.57	321	trace →
2	Muse Wonder Full	97.4%	TBA	405	trace →
3	Muse Wonder (xhigh)	96.6%	$69.06	490	trace →
4	Claude Fable 5.1 (xhigh)	96.1%	$72.31	309	trace →
5	GPT 6 Astra (max)	96.1%	$10.39	61	trace →
6	GPT-5.6 Sol (xhigh)	85.0%	$13.33	87	trace →
7	Muse Spark 1.3 (max)	76.6%	$8.94	274	trace →
8	Claude Opus 5 (xhigh)	75.5%	$77.19	277	trace →
9	GPT 5.5 (xhigh)	67.1%	$7.18	57	trace →
10	GPT 5.5 (high)	65.3%	$3.62	38	trace →
11	Claude Opus 4.8 (xhigh)	60.1%	$23.02	135	trace →
12	Muse Spark 1.2 (xhigh)	59.5%	$0.67	149	trace →
13	GPT 5.5	58.2%	$1.46	20	trace →
14	GPT-5.6 Sol	57.6%	$0.98	12	trace →
15	Gemini 3.5 Flash	57.4%	$9.42	157	trace →
16	Claude Opus 4.7 (xhigh)	56.3%	$16.22	187	trace →
17	Gemini 3.7 Flash	52.6%	$2.28	155	trace →
18	Claude Opus 4.6	52.6%	$16.83	237	trace →
19	Gemini 3.6 Flash	49.7%	$6.19	159	trace →
20	Claude Opus 4.7	47.4%	$2.46	58	trace →
21	GLM-5.2	46.6%	$37.55	312	trace →
22	Claude Sonnet 4.6	45.5%	$83.70	578	trace →
23	GPT 5.4	36.6%	$0.41	10	trace →
24	Muse Spark 1.3 (xhigh)	31.1%	$20.53	994	trace →
25	Gemini 3 Flash	28.9%	$0.46	95	trace →
26	GPT 5.4 mini	25.3%	$0.05	10	trace →
27	Claude Haiku 4.5	24.2%	$0.95	126	trace →
28	GPT 5 mini	23.2%	$0.03	20	trace →
29	Muse Spark 1.1 (xhigh)	4.7%	$0.80	182	trace →
30	Gemini 3.1 Pro	0.0%	$4.94	151	trace →

Click a row to replay how that model rebuilt this program, or the model name to open its full run