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

19 runs
# Model Score Cost Calls
1 GLM-5.2 98.9% $6.30 170 trace →
2 Gemini 3.6 Flash 98.6% $1.34 63 trace →
3 Claude Opus 4.6 98.2% $1.18 85 trace →
4 GPT 5.5 (high) 90.5% $1.93 28 trace →
5 GPT 5.6 Sol (xhigh) 87.7% $6.31 49 trace →
6 Claude Opus 4.8 (xhigh) 87.4% $5.31 63 trace →
7 Claude Opus 4.7 (xhigh) 87.0% $3.05 96 trace →
8 Gemini 3 Flash 86.7% $0.14 55 trace →
9 GPT 5.5 (xhigh) 82.5% $2.08 32 trace →
10 Gemini 3.5 Flash 82.5% $0.91 66 trace →
11 GPT 5.5 82.5% $0.85 15 trace →
12 GPT 5.4 80.4% $0.15 10 trace →
13 GPT 5.6 Sol (medium) 76.5% $0.53 8 trace →
14 Claude Opus 4.7 76.1% $0.77 43 trace →
15 Gemini 3.1 Pro 73.7% $0.77 48 trace →
16 Claude Sonnet 4.6 71.6% $1.13 100 trace →
17 Claude Haiku 4.5 54.7% $0.53 108 trace →
18 GPT 5.4 mini 46.7% $0.16 255 trace →
19 GPT 5 mini 1.4% $0.02 16 trace →

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