eradman/entr | ProgramBench

Run arbitrary commands when files change

586

Generated Behavioral Tests

99.8%

Best Score

Muse Wonder (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	Muse Wonder (xhigh)	99.8%	$16.35	243	trace →
2	Claude Opus 5.5 (xhigh)	99.7%	$15.52	228	trace →
3	Claude Fable 5.1 (xhigh)	99.5%	$24.17	177	trace →
4	GPT 6 Astra (max)	99.3%	$6.99	42	trace →
5	Claude Opus 5 (xhigh)	98.3%	$22.46	186	trace →
6	Muse Wonder Full	97.3%	TBA	343	trace →
7	GPT-5.6 Sol (xhigh)	95.6%	$3.62	27	trace →
8	Muse Spark 1.3 (max)	95.4%	$4.75	338	trace →
9	Muse Spark 1.3 (xhigh)	95.2%	$1.15	212	trace →
10	Gemini 3.7 Flash	93.5%	$0.75	92	trace →
11	GPT 5.5 (xhigh)	91.8%	$6.62	83	trace →
12	Claude Opus 4.8 (xhigh)	91.6%	$10.41	126	trace →
13	GLM-5.2	90.9%	$16.54	246	trace →
14	GPT 5.5 (high)	88.9%	$1.64	23	trace →
15	Claude Sonnet 4.6	88.6%	$12.14	383	trace →
16	Muse Spark 1.2 (xhigh)	84.3%	$0.80	271	trace →
17	Claude Opus 4.6	80.9%	$7.88	202	trace →
18	Claude Opus 4.7	79.7%	$1.89	77	trace →
19	GPT-5.6 Sol	79.2%	$0.74	15	trace →
20	Gemini 3.5 Flash	79.0%	$2.45	134	trace →
21	Gemini 3.6 Flash	78.0%	$2.06	108	trace →
22	GPT 5.5	75.9%	$0.86	19	trace →
23	Gemini 3.1 Pro	70.8%	$1.98	152	trace →
24	Gemini 3 Flash	65.7%	$0.25	77	trace →
25	GPT 5.4	58.5%	$0.24	13	trace →
26	Claude Opus 4.7 (xhigh)	42.8%	$33.60	454	trace →
27	Claude Haiku 4.5	35.5%	$0.53	109	trace →
28	GPT 5.4 mini	13.3%	$0.02	7	trace →
29	Muse Spark 1.1 (xhigh)	3.4%	$0.31	173	trace →
30	GPT 5 mini	1.2%	$0.01	12	trace →

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