iframe-proxy | Sunbelt Computer Software

Ping, but with a graph

338

Generated Behavioral Tests

97.3%

Best Score

Claude Opus 5 (xhigh)

Results by Model

Hover a point for details · The line marks the Pareto frontier (best score per cost) · Click a point to see model details

21 runs

#	Model	Score	Cost	Calls
1	Claude Opus 5 (xhigh)	97.3%	$58.88	294	trace →
2	Claude Opus 4.8 (xhigh)	95.0%	$20.74	134	trace →
3	Gemini 3.6 Flash	91.4%	$2.43	111	trace →
4	GLM-5.2	91.1%	$20.13	168	trace →
5	GPT 5.5 (xhigh)	88.5%	$8.13	76	trace →
6	GPT 5.5 (high)	85.5%	$3.77	49	trace →
7	GPT-5.6 Sol (xhigh)	84.9%	$4.00	25	trace →
8	GPT-5.6 Sol	80.5%	$1.47	23	trace →
9	Gemini 3.7 Flash	80.5%	$1.16	127	trace →
10	Claude Opus 4.6	79.0%	$10.87	211	trace →
11	Claude Opus 4.7 (xhigh)	77.5%	$5.92	120	trace →
12	Claude Opus 4.7	72.5%	$1.89	70	trace →
13	GPT 5.5	68.9%	$0.82	16	trace →
14	Claude Sonnet 4.6	58.3%	$11.38	271	trace →
15	Gemini 3.5 Flash	56.5%	$4.22	109	trace →
16	GPT 5.4	49.1%	$0.18	8	trace →
17	Gemini 3.1 Pro	37.0%	$0.66	33	trace →
18	Gemini 3 Flash	29.6%	$0.22	64	trace →
19	GPT 5 mini	21.9%	$0.01	7	trace →
20	GPT 5.4 mini	21.0%	$0.03	18	trace →
21	Claude Haiku 4.5	16.3%	$0.48	113	trace →

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