junegunn/fzf | ProgramBench

:cherry_blossom: A command-line fuzzy finder

1,856

Generated Behavioral Tests

98.9%

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)	98.9%	$75.12	537	trace →
2	Claude Fable 5.1 (xhigh)	97.3%	$60.97	263	trace →
3	Muse Wonder Full	96.6%	TBA	608	trace →
4	Claude Opus 5 (xhigh)	93.7%	$51.90	254	trace →
5	GPT 6 Astra (max)	93.4%	$14.46	65	trace →
6	Muse Wonder (xhigh)	89.9%	$50.64	463	trace →
7	Claude Opus 4.8 (xhigh)	84.9%	$27.12	139	trace →
8	GPT 5.5 (xhigh)	82.4%	$8.19	91	trace →
9	GPT-5.6 Sol (xhigh)	82.0%	$6.74	35	trace →
10	Muse Spark 1.2 (xhigh)	81.9%	$1.50	224	trace →
11	Claude Opus 4.6	81.9%	$11.78	253	trace →
12	Muse Spark 1.3 (xhigh)	81.4%	$1.33	244	trace →
13	Claude Sonnet 4.6	81.1%	$29.33	607	trace →
14	GPT 5.5 (high)	79.1%	$3.76	52	trace →
15	Muse Spark 1.3 (max)	79.0%	$3.72	290	trace →
16	GLM-5.2	77.9%	$8.45	143	trace →
17	GPT 5.5	77.5%	$0.97	17	trace →
18	Claude Opus 4.7 (xhigh)	77.2%	$1.80	53	trace →
19	Gemini 3.7 Flash	76.0%	$1.15	109	trace →
20	GPT-5.6 Sol	75.4%	$1.03	18	trace →
21	Claude Opus 4.7	74.8%	$0.56	20	trace →
22	GPT 5.4	71.9%	$0.34	12	trace →
23	Gemini 3.6 Flash	69.1%	$1.98	110	trace →
24	Gemini 3.5 Flash	68.7%	$3.32	120	trace →
25	Gemini 3.1 Pro	62.7%	$0.77	55	trace →
26	GPT 5.4 mini	46.6%	$0.04	7	trace →
27	Claude Haiku 4.5	22.3%	$0.72	105	trace →
28	Gemini 3 Flash	14.0%	$0.16	45	trace →
29	Muse Spark 1.1 (xhigh)	3.8%	$0.30	109	trace →
30	GPT 5 mini	3.6%	$0.03	12	trace →

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