ducaale/xh | ProgramBench

Friendly and fast tool for sending HTTP requests

1,171

Generated Behavioral Tests

96.2%

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)	96.2%	$70.95	446	trace →
2	Claude Opus 5 (xhigh)	84.8%	$109.61	482	trace →
3	GPT 6 Astra (max)	81.5%	$20.81	77	trace →
4	Claude Fable 5.1 (xhigh)	79.5%	$69.99	366	trace →
5	Muse Wonder (xhigh)	78.4%	$60.71	578	trace →
6	Muse Wonder Full	66.9%	TBA	457	trace →
7	GPT-5.6 Sol (xhigh)	61.5%	$9.94	73	trace →
8	Muse Spark 1.3 (xhigh)	60.6%	$1.09	188	trace →
9	Gemini 3.7 Flash	55.1%	$1.56	154	trace →
10	Muse Spark 1.1 (xhigh)	52.0%	$0.43	262	trace →
11	Claude Opus 4.6	50.0%	$13.67	284	trace →
12	GPT 5.5 (high)	48.5%	$2.92	27	trace →
13	GPT-5.6 Sol	45.9%	$1.17	17	trace →
14	Gemini 3.5 Flash	45.9%	$4.74	187	trace →
15	Gemini 3.6 Flash	44.3%	$3.04	113	trace →
16	GPT 5.5 (xhigh)	43.5%	$10.20	89	trace →
17	GPT 5.5	42.2%	$1.18	17	trace →
18	Claude Opus 4.7 (xhigh)	38.6%	$3.96	82	trace →
19	Claude Opus 4.8 (xhigh)	37.2%	$45.04	227	trace →
20	GPT 5.4	36.6%	$0.25	7	trace →
21	GLM-5.2	36.5%	$40.00	318	trace →
22	Claude Sonnet 4.6	33.6%	$45.79	702	trace →
23	Gemini 3.1 Pro	24.7%	$0.84	54	trace →
24	Muse Spark 1.2 (xhigh)	23.9%	$2.84	311	trace →
25	Claude Haiku 4.5	19.8%	$1.42	156	trace →
26	Gemini 3 Flash	16.7%	$0.23	73	trace →
27	Claude Opus 4.7	15.5%	$1.24	43	trace →
28	GPT 5 mini	9.4%	$0.02	13	trace →
29	Muse Spark 1.3 (max)	1.3%	$10.31	366	trace →
30	GPT 5.4 mini	1.3%	$0.07	14	trace →

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