sayanarijit/xplr | ProgramBench

A hackable, minimal, fast TUI file explorer

419

Generated Behavioral Tests

99.0%

Best Score

Claude Fable 5.1 (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 Fable 5.1 (xhigh)	99.0%	$100.65	442	trace →
2	Claude Opus 5.5 (xhigh)	98.3%	$76.05	409	trace →
3	Muse Wonder (xhigh)	97.9%	$53.35	508	trace →
4	Claude Opus 5 (xhigh)	97.1%	$78.47	344	trace →
5	GPT-5.6 Sol (xhigh)	81.6%	$7.83	42	trace →
6	Muse Spark 1.3 (max)	79.0%	$4.61	258	trace →
7	GPT 5.5 (xhigh)	78.5%	$9.61	90	trace →
8	Claude Opus 4.8 (xhigh)	77.1%	$29.21	206	trace →
9	Claude Opus 4.6	75.4%	$16.61	277	trace →
10	Muse Spark 1.3 (xhigh)	70.9%	$1.19	203	trace →
11	Gemini 3.7 Flash	70.6%	$1.91	152	trace →
12	Gemini 3.6 Flash	68.5%	$3.44	138	trace →
13	GPT 5.5 (high)	68.3%	$3.51	37	trace →
14	Muse Spark 1.2 (xhigh)	67.5%	$1.71	236	trace →
15	Muse Spark 1.1 (xhigh)	66.6%	$0.26	253	trace →
16	Claude Opus 4.7	66.6%	$4.36	108	trace →
17	Claude Opus 4.7 (xhigh)	63.0%	$2.74	65	trace →
18	Muse Wonder Full	44.6%	TBA	514	trace →
19	GPT 6 Astra (max)	44.6%	$19.97	80	trace →
20	Gemini 3.1 Pro	43.0%	$1.76	87	trace →
21	GLM-5.2	41.1%	$15.49	142	trace →
22	GPT 5.4	41.1%	$0.26	11	trace →
23	GPT-5.6 Sol	40.1%	$1.26	13	trace →
24	Claude Haiku 4.5	40.1%	$0.61	123	trace →
25	Gemini 3.5 Flash	39.6%	$4.19	154	trace →
26	Gemini 3 Flash	38.9%	$0.21	65	trace →
27	GPT 5.5	36.8%	$1.47	22	trace →
28	GPT 5.4 mini	35.8%	$0.02	7	trace →
29	Claude Sonnet 4.6	15.3%	$19.67	407	trace →
30	GPT 5 mini	14.8%	$0.02	12	trace →

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