google/brotli | ProgramBench

Brotli compression format

441

Generated Behavioral Tests

99.8%

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	99.8%	TBA	214	trace →
2	Claude Opus 5.5 (xhigh)	99.5%	$53.19	244	trace →
3	Muse Wonder (xhigh)	99.5%	$22.28	265	trace →
4	GPT 6 Astra (max)	96.6%	$18.72	91	trace →
5	GPT 5.5 (xhigh)	95.5%	$14.46	116	trace →
6	Muse Spark 1.3 (max)	95.2%	$4.45	258	trace →
7	Muse Spark 1.3 (xhigh)	93.7%	$1.16	210	trace →
8	Claude Sonnet 4.6	90.7%	$19.93	394	trace →
9	GLM-5.2	85.9%	$9.68	135	trace →
10	Claude Opus 4.7	84.6%	$5.68	110	trace →
11	GPT 5.5 (high)	82.1%	$4.86	59	trace →
12	Muse Spark 1.1 (xhigh)	82.1%	$3.19	247	trace →
13	Claude Opus 4.6	80.3%	$10.59	262	trace →
14	Gemini 3.7 Flash	79.6%	$1.69	121	trace →
15	Muse Spark 1.2 (xhigh)	78.9%	$0.40	264	trace →
16	Gemini 3.5 Flash	69.6%	$6.04	199	trace →
17	GPT-5.6 Sol	68.7%	$0.52	12	trace →
18	Gemini 3.6 Flash	61.7%	$3.46	133	trace →
19	Gemini 3.1 Pro	52.8%	$1.03	69	trace →
20	Gemini 3 Flash	46.7%	$0.34	100	trace →
21	GPT 5 mini	1.6%	$0.02	14	trace →
22	Claude Opus 5 (xhigh)	1.1%	$49.72	221	trace →
23	GPT 5.5	1.1%	$1.13	22	trace →
24	GPT 5.4	0.9%	$0.52	18	trace →
25	Claude Haiku 4.5	0.9%	$0.63	122	trace →
26	GPT-5.6 Sol (xhigh)	0.7%	$7.31	43	trace →
27	Claude Opus 4.8 (xhigh)	0.7%	$28.02	173	trace →
28	GPT 5.4 mini	0.7%	$0.05	11	trace →
29	Claude Opus 4.7 (xhigh)	0.5%	$6.32	94	trace →
30	Claude Fable 5.1 (xhigh)	0.0%	$59.72	127	trace →

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