sharkdp/pastel | ProgramBench

A command-line tool to generate, analyze, convert and manipulate colors

1,114

Generated Behavioral Tests

99.8%

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)	99.8%	$94.90	404	trace →
2	Muse Wonder Full	99.5%	TBA	377	trace →
3	Claude Opus 5.5 (xhigh)	99.4%	$36.90	355	trace →
4	GPT 6 Astra (max)	99.3%	$9.20	56	trace →
5	Muse Wonder (xhigh)	99.1%	$46.98	382	trace →
6	Claude Opus 5 (xhigh)	98.7%	$87.06	352	trace →
7	Muse Spark 1.3 (max)	95.4%	$6.42	353	trace →
8	Claude Opus 4.8 (xhigh)	94.7%	$30.75	181	trace →
9	Muse Spark 1.3 (xhigh)	94.3%	$3.66	400	trace →
10	GPT-5.6 Sol (xhigh)	93.4%	$8.76	69	trace →
11	GPT 5.5 (xhigh)	93.1%	$10.61	87	trace →
12	Muse Spark 1.2 (xhigh)	91.1%	$3.90	405	trace →
13	Claude Opus 4.7 (xhigh)	89.7%	$30.49	286	trace →
14	GPT 5.5 (high)	85.3%	$5.06	48	trace →
15	Gemini 3.6 Flash	84.1%	$5.17	139	trace →
16	Muse Spark 1.1 (xhigh)	84.0%	$1.28	224	trace →
17	GLM-5.2	83.8%	$32.12	222	trace →
18	Gemini 3.5 Flash	82.3%	$10.77	199	trace →
19	Gemini 3.7 Flash	81.0%	$5.76	195	trace →
20	GPT 5.5	77.8%	$1.84	21	trace →
21	Claude Opus 4.6	77.2%	$17.50	232	trace →
22	GPT-5.6 Sol	75.3%	$1.57	27	trace →
23	Claude Opus 4.7	74.6%	$3.22	92	trace →
24	GPT 5.4	50.6%	$0.22	7	trace →
25	Claude Haiku 4.5	35.4%	$1.22	179	trace →
26	Gemini 3.1 Pro	27.0%	$1.62	73	trace →
27	GPT 5 mini	18.7%	$0.04	32	trace →
28	Gemini 3 Flash	17.8%	$0.38	82	trace →
29	GPT 5.4 mini	9.4%	$0.04	21	trace →
30	Claude Sonnet 4.6	0.0%	$6.72	321	trace →

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