go-critic/go-critic | ProgramBench

The most opinionated Go source code linter for code audit.

493

Generated Behavioral Tests

81.7%

Best Score

Muse Wonder Full

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	Muse Wonder Full	81.7%	TBA	733	trace →
2	Claude Opus 5.5 (xhigh)	81.3%	$88.95	665	trace →
3	GPT 6 Astra (max)	78.3%	$12.33	54	trace →
4	Muse Wonder (xhigh)	73.4%	$60.38	616	trace →
5	GPT 5.5 (xhigh)	56.8%	$9.11	82	trace →
6	Muse Spark 1.3 (max)	56.0%	$3.97	310	trace →
7	Muse Spark 1.3 (xhigh)	55.6%	$1.26	218	trace →
8	Gemini 3.7 Flash	55.6%	$1.53	105	trace →
9	GPT-5.6 Sol (xhigh)	53.8%	$4.53	38	trace →
10	Claude Opus 4.8 (xhigh)	52.7%	$32.98	209	trace →
11	GPT 5.5 (high)	49.7%	$3.64	43	trace →
12	Gemini 3.6 Flash	46.7%	$2.19	116	trace →
13	Gemini 3.5 Flash	45.2%	$4.14	171	trace →
14	Muse Spark 1.1 (xhigh)	42.2%	$3.14	264	trace →
15	Claude Sonnet 4.6	41.6%	$34.12	677	trace →
16	GLM-5.2	39.4%	$35.46	321	trace →
17	Claude Opus 4.6	35.1%	$20.28	360	trace →
18	GPT 5.5	34.1%	$1.19	21	trace →
19	GPT-5.6 Sol	31.4%	$0.65	11	trace →
20	GPT 5.4	30.4%	$0.49	14	trace →
21	Claude Opus 4.7	28.2%	$4.00	134	trace →
22	Claude Opus 4.7 (xhigh)	27.8%	$4.79	122	trace →
23	Gemini 3 Flash	26.0%	$0.21	58	trace →
24	Gemini 3.1 Pro	23.9%	$1.15	111	trace →
25	Claude Opus 5 (xhigh)	22.3%	$16.62	141	trace →
26	Claude Haiku 4.5	21.7%	$0.46	89	trace →
27	Claude Fable 5.1 (xhigh)	13.8%	$6.20	88	trace →
28	Muse Spark 1.2 (xhigh)	8.3%	$0.18	100	trace →
29	GPT 5.4 mini	3.9%	$0.04	7	trace →
30	GPT 5 mini	0.6%	$0.02	20	trace →

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