← Back to leaderboard  ·  Show all task instances

Miserlou/Loop

UNIX's missing `loop` command

692 rs easy
710
Generated Behavioral Tests
99.3%
Best Score
Claude Opus 5 (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)

Per-Model Results

25 runs
# Model Score Cost Calls
1 Claude Opus 5 (xhigh) 99.3% $26.22 195 trace →
2 Claude Opus 4.8 (xhigh) 98.9% $18.57 135 trace →
3 GLM-5.2 98.1% $17.72 280 trace →
4 GPT 5.5 (xhigh) 97.3% $4.91 66 trace →
5 Muse Spark 1.3 (max) 96.6% $3.57 320 trace →
6 Muse Spark 1.3 (xhigh) 95.5% $0.67 162 trace →
7 Muse Spark 1.1 (xhigh) 94.8% $1.39 225 trace →
8 Claude Opus 4.6 94.6% $8.74 301 trace →
9 Claude Opus 4.7 (xhigh) 94.4% $10.93 196 trace →
10 Muse Spark 1.2 (xhigh) 93.8% $1.37 189 trace →
11 Gemini 3.7 Flash 93.4% $1.36 140 trace →
12 Gemini 3.6 Flash 92.0% $4.66 223 trace →
13 Gemini 3.5 Flash 91.3% $4.35 174 trace →
14 GPT-5.6 Sol (xhigh) 83.9% $1.73 18 trace →
15 GPT 5.5 (high) 83.0% $4.57 50 trace →
16 Gemini 3.1 Pro 81.7% $1.66 176 trace →
17 Claude Opus 4.7 79.3% $5.22 143 trace →
18 GPT 5.5 55.4% $1.07 18 trace →
19 Claude Haiku 4.5 54.9% $0.94 145 trace →
20 GPT-5.6 Sol 53.0% $0.29 9 trace →
21 GPT 5.4 52.0% $0.24 12 trace →
22 Gemini 3 Flash 52.0% $0.51 142 trace →
23 Claude Sonnet 4.6 45.8% $7.56 370 trace →
24 GPT 5 mini 24.2% $0.02 15 trace →
25 GPT 5.4 mini 0.0% $0.04 8 trace →

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