htop-dev/htop | ProgramBench

htop - an interactive process viewer

692

Generated Behavioral Tests

99.4%

Best Score

GPT-5.6 Sol (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)

25 runs

#	Model	Score	Cost	Calls
1	GPT-5.6 Sol (xhigh)	99.4%	$7.51	38	trace →
2	Claude Opus 4.8 (xhigh)	98.5%	$10.49	93	trace →
3	GLM-5.2	98.3%	$10.83	142	trace →
4	GPT 5.5 (xhigh)	98.1%	$4.18	47	trace →
5	Muse Spark 1.3 (max)	97.1%	$2.56	148	trace →
6	Muse Spark 1.3 (xhigh)	96.4%	$0.43	123	trace →
7	Gemini 3.6 Flash	96.2%	$3.14	106	trace →
8	Gemini 3.5 Flash	95.2%	$2.89	84	trace →
9	GPT-5.6 Sol	87.4%	$0.64	11	trace →
10	Claude Opus 4.6	85.3%	$14.15	238	trace →
11	GPT 5.5 (high)	84.8%	$2.63	28	trace →
12	Gemini 3.7 Flash	83.5%	$1.68	88	trace →
13	Gemini 3.1 Pro	80.2%	$0.93	55	trace →
14	GPT 5.5	75.7%	$0.80	13	trace →
15	GPT 5.4	15.3%	$0.28	8	trace →
16	Muse Spark 1.1 (xhigh)	10.3%	$0.38	87	trace →
17	Muse Spark 1.2 (xhigh)	10.1%	$0.11	105	trace →
18	Claude Opus 4.7 (xhigh)	9.8%	$3.13	104	trace →
19	Claude Opus 4.7	9.8%	$1.55	59	trace →
20	GPT 5.4 mini	9.0%	$0.03	10	trace →
21	Claude Sonnet 4.6	7.2%	$11.57	259	trace →
22	Claude Haiku 4.5	7.2%	$0.56	107	trace →
23	GPT 5 mini	6.6%	$0.02	13	trace →
24	Gemini 3 Flash	6.5%	$0.21	43	trace →
25	Claude Opus 5 (xhigh)	0.0%	$5.52	73	trace →

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