ajeetdsouza/zoxide | ProgramBench

A smarter cd command. Supports all major shells.

531

Generated Behavioral Tests

99.2%

Best Score

GPT 6 Astra (max)

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	GPT 6 Astra (max)	99.2%	$10.23	53	trace →
2	Claude Opus 5.5 (xhigh)	98.9%	$8.87	179	trace →
3	Muse Wonder (xhigh)	98.3%	$18.92	251	trace →
4	Claude Fable 5.1 (xhigh)	98.1%	$22.96	175	trace →
5	Muse Wonder Full	98.1%	TBA	215	trace →
6	Claude Opus 5 (xhigh)	97.9%	$83.64	311	trace →
7	Muse Spark 1.3 (max)	97.6%	$4.34	317	trace →
8	Muse Spark 1.3 (xhigh)	95.9%	$1.54	246	trace →
9	GPT-5.6 Sol (xhigh)	90.2%	$6.36	42	trace →
10	Claude Opus 4.7 (xhigh)	87.2%	$32.86	378	trace →
11	GLM-5.2	86.1%	$61.93	312	trace →
12	Claude Opus 4.8 (xhigh)	85.1%	$31.19	170	trace →
13	Gemini 3.7 Flash	84.4%	$1.71	135	trace →
14	GPT 5.5 (xhigh)	83.4%	$10.70	98	trace →
15	Gemini 3.5 Flash	81.5%	$5.28	204	trace →
16	GPT 5.5 (high)	79.8%	$4.39	47	trace →
17	Gemini 3.6 Flash	78.5%	$3.18	144	trace →
18	Claude Opus 4.7	76.5%	$10.07	187	trace →
19	Muse Spark 1.1 (xhigh)	75.7%	$0.32	243	trace →
20	Claude Sonnet 4.6	69.9%	$14.61	368	trace →
21	Muse Spark 1.2 (xhigh)	69.7%	$0.20	238	trace →
22	GPT 5.5	65.0%	$0.95	16	trace →
23	Claude Opus 4.6	63.7%	$18.43	307	trace →
24	Gemini 3.1 Pro	58.2%	$2.10	121	trace →
25	GPT 5.4	53.5%	$0.34	9	trace →
26	Gemini 3 Flash	52.0%	$0.42	157	trace →
27	Claude Haiku 4.5	46.7%	$0.88	112	trace →
28	GPT 5.4 mini	1.5%	$0.03	8	trace →
29	GPT-5.6 Sol	1.1%	$1.52	21	trace →
30	GPT 5 mini	0.0%	$0.05	15	trace →

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