kaushiksrini/parqeye

Peek inside Parquet files right from your terminal

654 rs medium
479
Generated Behavioral Tests
99.0%
Best Score
Claude Opus 5 (xhigh)

Hover a point for details · The line marks the Pareto frontier (best score per cost) · Click a point to see model details

20 runs
# Model Score Cost Calls
1 Claude Opus 5 (xhigh) 99.0% $155.70 486 trace →
2 GPT 5.5 (xhigh) 81.6% $6.37 55 trace →
3 Claude Opus 4.8 (xhigh) 73.4% $23.62 150 trace →
4 GPT-5.6 Sol (xhigh) 71.8% $11.63 76 trace →
5 GLM-5.2 69.9% $74.17 315 trace →
6 Gemini 3.6 Flash 69.1% $5.24 138 trace →
7 GPT-5.6 Sol 68.1% $1.40 17 trace →
8 GPT 5.5 (high) 64.1% $4.43 39 trace →
9 Claude Opus 4.7 (xhigh) 64.1% $4.31 112 trace →
10 GPT 5.5 60.3% $1.66 24 trace →
11 Claude Opus 4.7 58.9% $1.25 40 trace →
12 GPT 5.4 48.4% $0.21 14 trace →
13 Claude Opus 4.6 45.5% $14.19 203 trace →
14 Claude Haiku 4.5 34.4% $1.16 181 trace →
15 GPT 5 mini 13.8% $0.02 17 trace →
16 Gemini 3.5 Flash 13.2% $17.60 262 trace →
17 Claude Sonnet 4.6 13.2% $91.59 720 trace →
18 Gemini 3.1 Pro 4.2% $0.82 46 trace →
19 Gemini 3 Flash 0.0% $0.41 105 trace →
20 GPT 5.4 mini 0.0% $0.02 18 trace →

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