nikolassv/bartib | ProgramBench

A simple timetracker for the command line. It saves a log of all tracked activities as a plaintext file and allows you to create flexible reports.

721

Generated Behavioral Tests

98.9%

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	98.9%	TBA	256	trace →
2	Claude Opus 5.5 (xhigh)	98.1%	$25.39	237	trace →
3	Muse Wonder (xhigh)	97.9%	$32.99	313	trace →
4	GPT 6 Astra (max)	97.6%	$12.46	70	trace →
5	Muse Spark 1.3 (max)	97.1%	$3.15	194	trace →
6	Claude Opus 5 (xhigh)	96.9%	$18.41	146	trace →
7	Claude Fable 5.1 (xhigh)	96.7%	$25.69	194	trace →
8	Muse Spark 1.3 (xhigh)	96.5%	$1.89	236	trace →
9	GPT 5.5 (xhigh)	96.0%	$9.37	89	trace →
10	GPT-5.6 Sol (xhigh)	93.9%	$4.21	21	trace →
11	GLM-5.2	93.9%	$25.70	288	trace →
12	Gemini 3.7 Flash	91.7%	$3.02	126	trace →
13	GPT 5.5 (high)	90.4%	$4.12	39	trace →
14	Gemini 3.6 Flash	88.2%	$4.73	141	trace →
15	Claude Opus 4.7	87.2%	$13.55	219	trace →
16	Claude Opus 4.7 (xhigh)	84.6%	$17.84	250	trace →
17	Muse Spark 1.2 (xhigh)	84.5%	$0.80	133	trace →
18	Claude Sonnet 4.6	84.3%	$13.63	283	trace →
19	Claude Opus 4.8 (xhigh)	84.3%	$33.86	204	trace →
20	GPT-5.6 Sol	82.5%	$1.00	15	trace →
21	Claude Opus 4.6	80.9%	$10.92	274	trace →
22	Gemini 3.5 Flash	74.8%	$7.01	195	trace →
23	GPT 5.5	64.6%	$0.94	14	trace →
24	Gemini 3.1 Pro	62.4%	$1.58	99	trace →
25	GPT 5.4	58.7%	$0.42	12	trace →
26	Gemini 3 Flash	51.2%	$0.25	86	trace →
27	Claude Haiku 4.5	48.4%	$0.89	149	trace →
28	GPT 5.4 mini	23.3%	$0.06	10	trace →
29	GPT 5 mini	18.7%	$0.04	19	trace →
30	Muse Spark 1.1 (xhigh)	2.8%	$0.58	121	trace →

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