codesnap-rs/codesnap | ProgramBench

🦀️📸 Pure Rust tool to generate beautiful code snapshots, provide CLI and Library

571

Generated Behavioral Tests

92.8%

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	92.8%	TBA	347	trace →
2	Muse Wonder (xhigh)	92.5%	$23.49	299	trace →
3	Claude Opus 5 (xhigh)	91.9%	$46.91	322	trace →
4	Muse Spark 1.3 (max)	91.9%	$5.24	405	trace →
5	GLM-5.2	88.6%	$22.86	179	trace →
6	Muse Spark 1.3 (xhigh)	87.4%	$0.54	155	trace →
7	Muse Spark 1.1 (xhigh)	83.9%	$0.10	108	trace →
8	Claude Opus 4.8 (xhigh)	83.3%	$14.74	135	trace →
9	GPT 5.5 (xhigh)	82.7%	$7.25	63	trace →
10	Gemini 3.6 Flash	82.0%	$4.02	164	trace →
11	GPT 5.5 (high)	79.0%	$3.04	34	trace →
12	Muse Spark 1.2 (xhigh)	78.6%	$0.41	110	trace →
13	GPT 5.4	75.7%	$0.79	11	trace →
14	Gemini 3.1 Pro	69.7%	$1.40	83	trace →
15	Claude Opus 5.5 (xhigh)	69.5%	$42.78	416	trace →
16	Claude Fable 5.1 (xhigh)	69.0%	$47.85	287	trace →
17	GPT 6 Astra (max)	69.0%	$9.69	64	trace →
18	GPT-5.6 Sol	68.0%	$0.67	8	trace →
19	GPT 5.4 mini	62.5%	$0.08	41	trace →
20	GPT-5.6 Sol (xhigh)	62.2%	$4.71	37	trace →
21	Gemini 3.5 Flash	62.0%	$6.80	230	trace →
22	Claude Opus 4.7	60.2%	$1.88	54	trace →
23	GPT 5.5	59.4%	$0.69	15	trace →
24	Claude Haiku 4.5	51.0%	$0.49	86	trace →
25	Gemini 3 Flash	29.2%	$0.51	116	trace →
26	GPT 5 mini	11.0%	$0.01	9	trace →
27	Gemini 3.7 Flash	9.5%	$1.11	124	trace →
28	Claude Opus 4.6	9.5%	$8.54	172	trace →
29	Claude Sonnet 4.6	9.5%	$20.92	395	trace →
30	Claude Opus 4.7 (xhigh)	7.5%	$4.86	120	trace →

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