boyter/scc | ProgramBench

Sloc, Cloc and Code: scc is a very fast accurate code counter with complexity calculations and COCOMO estimates written in pure Go

464

Generated Behavioral Tests

88.6%

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	88.6%	TBA	506	trace →
2	Claude Opus 5.5 (xhigh)	86.6%	$51.40	449	trace →
3	GPT 6 Astra (max)	73.7%	$14.15	81	trace →
4	Claude Opus 5 (xhigh)	72.8%	$87.44	346	trace →
5	Muse Wonder (xhigh)	72.4%	$68.62	581	trace →
6	Muse Spark 1.3 (xhigh)	57.5%	$1.73	235	trace →
7	Gemini 3.7 Flash	49.4%	$1.58	118	trace →
8	Claude Fable 5.1 (xhigh)	48.3%	$99.11	342	trace →
9	GPT 5.5 (high)	47.6%	$5.71	67	trace →
10	Gemini 3.5 Flash	40.5%	$6.55	177	trace →
11	Muse Spark 1.2 (xhigh)	37.7%	$1.90	337	trace →
12	Claude Sonnet 4.6	37.7%	$46.35	593	trace →
13	Claude Opus 4.6	37.1%	$15.01	315	trace →
14	Claude Opus 4.7 (xhigh)	36.4%	$20.65	267	trace →
15	GLM-5.2	34.3%	$31.59	223	trace →
16	Muse Spark 1.1 (xhigh)	34.1%	$0.49	219	trace →
17	Gemini 3.6 Flash	32.1%	$4.31	141	trace →
18	Claude Opus 4.8 (xhigh)	31.0%	$14.64	125	trace →
19	GPT 5.5 (xhigh)	28.2%	$17.09	107	trace →
20	GPT-5.6 Sol	28.2%	$1.33	20	trace →
21	GPT 5.5	20.7%	$0.83	11	trace →
22	Claude Opus 4.7	19.6%	$1.69	54	trace →
23	GPT-5.6 Sol (xhigh)	18.5%	$10.39	65	trace →
24	Muse Spark 1.3 (max)	12.7%	$5.83	351	trace →
25	Gemini 3.1 Pro	11.6%	$1.22	98	trace →
26	Claude Haiku 4.5	8.2%	$0.60	86	trace →
27	Gemini 3 Flash	5.8%	$0.23	76	trace →
28	GPT 5.4	4.5%	$0.26	9	trace →
29	GPT 5 mini	1.9%	$0.03	13	trace →
30	GPT 5.4 mini	1.7%	$0.03	9	trace →

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