universal-ctags/ctags | ProgramBench

A maintained ctags implementation

2,258

Generated Behavioral Tests

41.8%

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)	41.8%	$36.85	320	trace →
2	Muse Wonder (xhigh)	34.9%	$53.64	559	trace →
3	GPT 6 Astra (max)	33.8%	$13.50	67	trace →
4	GPT-5.6 Sol (xhigh)	19.0%	$7.27	45	trace →
5	Muse Spark 1.3 (max)	15.3%	$9.86	313	trace →
6	Claude Fable 5.1 (xhigh)	14.2%	$67.21	220	trace →
7	Muse Wonder Full	13.5%	TBA	178	trace →
8	Claude Sonnet 4.6	13.3%	$58.28	659	trace →
9	Muse Spark 1.3 (xhigh)	11.6%	$1.71	237	trace →
10	Claude Opus 4.8 (xhigh)	10.6%	$31.11	206	trace →
11	Muse Spark 1.2 (xhigh)	10.2%	$1.03	136	trace →
12	GPT-5.6 Sol	9.5%	$1.75	16	trace →
13	GPT 5.5	9.3%	$0.92	13	trace →
14	Claude Opus 4.7 (xhigh)	8.2%	$14.35	173	trace →
15	Gemini 3.5 Flash	7.2%	$4.40	111	trace →
16	GLM-5.2	4.7%	$25.49	244	trace →
17	GPT 5.5 (high)	4.4%	$3.65	31	trace →
18	GPT 5.4	3.6%	$0.20	8	trace →
19	Gemini 3 Flash	1.7%	$0.27	58	trace →
20	Gemini 3.1 Pro	1.6%	$1.44	118	trace →
21	Claude Haiku 4.5	1.0%	$0.71	110	trace →
22	GPT 5 mini	0.9%	$0.03	16	trace →
23	Claude Opus 4.7	0.5%	$0.29	22	trace →
24	GPT 5.4 mini	0.4%	$0.03	16	trace →
25	Gemini 3.6 Flash	0.1%	$3.43	155	trace →
26	Gemini 3.7 Flash	0.1%	$2.17	164	trace →
27	Muse Spark 1.1 (xhigh)	0.1%	$0.18	153	trace →
28	Claude Opus 5 (xhigh)	0.0%	$4.37	75	trace →
29	GPT 5.5 (xhigh)	0.0%	$3.75	34	trace →
30	Claude Opus 4.6	0.0%	$2.49	226	trace →

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