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 Claude Opus 4.8 (xhigh) 76.0% $27.10 158 trace →
2 GPT 5.6 Sol (xhigh) 70.3% $5.79 39 trace →
3 Gemini 3.6 Flash 65.1% $6.50 183 trace →
4 Claude Sonnet 4.6 63.5% $43.00 547 trace →
5 Claude Opus 4.6 60.9% $13.80 210 trace →
6 GPT 5.5 (high) 60.5% $6.36 60 trace →
7 GLM-5.2 56.2% $37.63 246 trace →
8 GPT 5.5 (xhigh) 54.9% $7.10 66 trace →
9 GPT 5.6 Sol (medium) 49.7% $1.50 22 trace →
10 Gemini 3.1 Pro 49.3% $1.41 100 trace →
11 GPT 5.5 38.5% $1.07 15 trace →
12 Gemini 3.5 Flash 14.8% $3.96 115 trace →
13 Claude Opus 4.7 13.9% $1.05 37 trace →
14 Claude Haiku 4.5 8.9% $0.91 156 trace →
15 GPT 5.4 8.6% $0.47 14 trace →
16 Gemini 3 Flash 3.1% $0.22 66 trace →
17 GPT 5.4 mini 0.3% $0.04 8 trace →
18 GPT 5 mini 0.1% $0.02 13 trace →
19 Claude Opus 4.7 (xhigh) 0.0% $4.35 65 trace →

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