ivanceras/svgbob | ProgramBench

Convert your ascii diagram scribbles into happy little SVG

472

Generated Behavioral Tests

96.4%

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)	96.4%	$49.62	408	trace →
2	GPT 6 Astra (max)	89.0%	$10.19	46	trace →
3	Muse Wonder Full	81.8%	TBA	339	trace →
4	Claude Fable 5.1 (xhigh)	77.5%	$75.43	266	trace →
5	Muse Wonder (xhigh)	73.3%	$35.28	291	trace →
6	Claude Opus 5 (xhigh)	64.6%	$35.35	168	trace →
7	Muse Spark 1.3 (xhigh)	55.3%	$2.76	292	trace →
8	Claude Opus 4.8 (xhigh)	54.9%	$17.37	139	trace →
9	GPT-5.6 Sol (xhigh)	53.6%	$6.66	37	trace →
10	Muse Spark 1.3 (max)	53.2%	$4.63	248	trace →
11	Gemini 3.7 Flash	47.9%	$3.29	118	trace →
12	GPT 5.5 (xhigh)	47.7%	$5.84	56	trace →
13	GPT 5.5 (high)	46.0%	$3.58	40	trace →
14	GLM-5.2	43.9%	$28.49	215	trace →
15	Claude Sonnet 4.6	41.3%	$43.75	744	trace →
16	Muse Spark 1.2 (xhigh)	41.1%	$0.38	160	trace →
17	Gemini 3.6 Flash	39.8%	$13.38	229	trace →
18	Gemini 3.5 Flash	39.8%	$5.65	147	trace →
19	GPT-5.6 Sol	39.2%	$0.78	12	trace →
20	GPT 5.5	35.2%	$1.00	17	trace →
21	Claude Opus 4.7 (xhigh)	31.6%	$4.11	95	trace →
22	Gemini 3.1 Pro	24.8%	$0.97	38	trace →
23	Claude Opus 4.7	23.1%	$0.75	36	trace →
24	Claude Opus 4.6	23.1%	$14.16	268	trace →
25	Gemini 3 Flash	21.8%	$0.27	52	trace →
26	GPT 5.4	18.6%	$0.15	6	trace →
27	Claude Haiku 4.5	18.0%	$1.28	187	trace →
28	GPT 5.4 mini	13.3%	$0.09	23	trace →
29	GPT 5 mini	2.3%	$0.03	16	trace →
30	Muse Spark 1.1 (xhigh)	0.0%	$0.68	124	trace →

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