mfridman/tparse | ProgramBench

CLI tool for summarizing go test output. Pipe friendly. CI/CD friendly.

425

Generated Behavioral Tests

99.3%

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)	99.3%	$18.87	236	trace →
2	Muse Wonder (xhigh)	95.5%	$41.69	411	trace →
3	Muse Wonder Full	95.3%	TBA	194	trace →
4	Claude Fable 5.1 (xhigh)	94.4%	$26.54	195	trace →
5	GPT 6 Astra (max)	93.2%	$12.00	63	trace →
6	Claude Opus 5 (xhigh)	91.8%	$36.01	222	trace →
7	GPT-5.6 Sol (xhigh)	90.8%	$8.73	61	trace →
8	Muse Spark 1.3 (xhigh)	90.4%	$1.47	204	trace →
9	Gemini 3.6 Flash	89.2%	$5.21	170	trace →
10	Gemini 3.7 Flash	85.9%	$2.28	148	trace →
11	Muse Spark 1.3 (max)	84.9%	$4.62	262	trace →
12	GLM-5.2	83.1%	$25.34	170	trace →
13	GPT 5.5 (high)	82.4%	$4.88	48	trace →
14	GPT 5.5 (xhigh)	81.6%	$7.81	81	trace →
15	Muse Spark 1.2 (xhigh)	81.6%	$4.27	257	trace →
16	Gemini 3.5 Flash	80.7%	$5.45	176	trace →
17	Claude Opus 4.7 (xhigh)	80.2%	$15.52	168	trace →
18	Claude Opus 4.6	77.6%	$13.53	179	trace →
19	Muse Spark 1.1 (xhigh)	76.9%	$2.01	289	trace →
20	Claude Sonnet 4.6	76.0%	$21.05	361	trace →
21	Claude Opus 4.8 (xhigh)	74.1%	$15.57	147	trace →
22	Claude Opus 4.7	72.0%	$2.33	69	trace →
23	GPT 5.5	70.1%	$1.50	23	trace →
24	GPT-5.6 Sol	68.9%	$0.54	11	trace →
25	Gemini 3 Flash	56.9%	$0.26	69	trace →
26	Gemini 3.1 Pro	51.8%	$1.05	55	trace →
27	Claude Haiku 4.5	51.5%	$0.62	84	trace →
28	GPT 5.4	39.1%	$0.24	8	trace →
29	GPT 5 mini	28.7%	$0.03	13	trace →
30	GPT 5.4 mini	6.1%	$0.02	6	trace →

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