← Back to leaderboard  ·  Show all task instances

dundee/gdu

Fast disk usage analyzer with console interface written in Go

5,578 go medium
1,161
Generated Behavioral Tests
89.8%
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) 89.8% $79.04 383 trace →
2 Muse Spark 1.3 (max) 86.6% $9.55 370 trace →
3 Muse Spark 1.2 (xhigh) 81.1% $1.29 315 trace →
4 GPT-5.6 Sol (xhigh) 80.0% $4.20 31 trace →
5 GPT 5.5 (xhigh) 78.9% $7.44 85 trace →
6 Muse Spark 1.3 (xhigh) 76.4% $1.38 249 trace →
7 GPT 5.5 (high) 75.8% $3.88 60 trace →
8 GPT-5.6 Sol 72.2% $1.11 16 trace →
9 Claude Sonnet 4.6 70.4% $21.13 532 trace →
10 Claude Opus 4.8 (xhigh) 70.3% $13.86 107 trace →
11 Claude Opus 4.6 66.6% $10.24 245 trace →
12 GLM-5.2 64.8% $23.74 274 trace →
13 GPT 5.5 64.3% $1.02 18 trace →
14 Gemini 3.7 Flash 63.5% $1.18 114 trace →
15 Claude Opus 4.7 (xhigh) 63.3% $5.45 96 trace →
16 Gemini 3.6 Flash 63.0% $5.15 177 trace →
17 Gemini 3.5 Flash 62.6% $5.06 163 trace →
18 Claude Opus 4.7 59.6% $2.35 69 trace →
19 Claude Haiku 4.5 42.1% $0.72 125 trace →
20 Gemini 3.1 Pro 36.3% $1.17 105 trace →
21 GPT 5.4 mini 32.4% $0.03 7 trace →
22 GPT 5.4 27.9% $0.31 8 trace →
23 Gemini 3 Flash 9.5% $0.28 73 trace →
24 GPT 5 mini 4.5% $0.01 6 trace →
25 Muse Spark 1.1 (xhigh) 2.6% $0.35 242 trace →

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