← Back to leaderboard  ·  Show all task instances

crowdagger/crowbook

Converts books written in Markdown to HTML, LaTeX/PDF and EPUB

813 rs medium
807
Generated Behavioral Tests
85.9%
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) 85.9% $118.23 503 trace →
2 Muse Spark 1.3 (xhigh) 67.2% $1.90 308 trace →
3 Gemini 3.7 Flash 66.5% $1.77 143 trace →
4 Claude Sonnet 4.6 60.3% $85.27 937 trace →
5 Claude Opus 4.8 (xhigh) 54.3% $24.10 188 trace →
6 Gemini 3.5 Flash 50.7% $7.78 244 trace →
7 GPT 5.5 (high) 48.8% $3.32 38 trace →
8 GPT-5.6 Sol 46.8% $1.24 16 trace →
9 Claude Opus 4.6 45.0% $19.00 320 trace →
10 Muse Spark 1.2 (xhigh) 44.4% $0.42 98 trace →
11 GPT 5.5 43.2% $1.04 15 trace →
12 Muse Spark 1.1 (xhigh) 40.0% $0.62 243 trace →
13 GPT-5.6 Sol (xhigh) 30.1% $10.67 73 trace →
14 Claude Opus 4.7 (xhigh) 29.4% $13.25 195 trace →
15 Gemini 3.1 Pro 22.4% $0.98 81 trace →
16 Claude Haiku 4.5 21.2% $0.47 106 trace →
17 GPT 5.4 20.6% $0.34 11 trace →
18 GPT 5 mini 13.6% $0.04 17 trace →
19 Gemini 3 Flash 12.8% $0.45 92 trace →
20 GPT 5.4 mini 11.5% $0.03 9 trace →
21 GPT 5.5 (xhigh) 2.9% $10.81 98 trace →
22 Gemini 3.6 Flash 2.9% $4.59 188 trace →
23 GLM-5.2 0.9% $21.10 230 trace →
24 Claude Opus 4.7 0.9% $2.01 69 trace →
25 Muse Spark 1.3 (max) 0.0% $7.00 291 trace →

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