← Back to leaderboard  ·  Show all task instances

zevv/duc

Dude, where are my bytes: Duc, a library and suite of tools for inspecting disk usage

682 c medium
874
Generated Behavioral Tests
95.9%
Best Score
Muse Spark 1.3 (max)

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 Muse Spark 1.3 (max) 95.9% $7.23 367 trace →
2 GPT 5.5 (xhigh) 93.0% $9.70 86 trace →
3 Muse Spark 1.3 (xhigh) 88.7% $2.78 351 trace →
4 Gemini 3.7 Flash 88.2% $2.43 176 trace →
5 Claude Opus 4.7 (xhigh) 87.5% $21.07 218 trace →
6 Muse Spark 1.1 (xhigh) 87.0% $1.57 433 trace →
7 GLM-5.2 84.3% $30.60 251 trace →
8 Claude Opus 4.7 83.4% $2.14 201 trace →
9 Gemini 3.6 Flash 81.9% $4.04 138 trace →
10 GPT 5.5 (high) 80.0% $3.81 40 trace →
11 GPT-5.6 Sol 79.6% $1.26 15 trace →
12 Gemini 3.5 Flash 78.9% $6.60 173 trace →
13 Claude Opus 4.8 (xhigh) 75.4% $32.62 171 trace →
14 Muse Spark 1.2 (xhigh) 72.2% $1.80 251 trace →
15 GPT 5.5 71.1% $1.19 17 trace →
16 Claude Opus 5 (xhigh) 57.8% $28.43 216 trace →
17 GPT 5.4 40.7% $0.30 7 trace →
18 GPT 5.4 mini 25.1% $0.06 7 trace →
19 Gemini 3 Flash 23.9% $0.40 78 trace →
20 GPT 5 mini 16.8% $0.02 13 trace →
21 Claude Haiku 4.5 16.7% $1.33 155 trace →
22 Claude Sonnet 4.6 11.8% $30.03 441 trace →
23 GPT-5.6 Sol (xhigh) 11.7% $7.92 45 trace →
24 Claude Opus 4.6 0.0% $5.00 228 trace →
25 Gemini 3.1 Pro 0.0% $2.24 142 trace →

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