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 GPT 5.6 Sol (xhigh) 63.1% $12.01 80 trace →
2 GPT 5.5 (xhigh) 55.0% $8.05 68 trace →
3 Claude Opus 4.6 53.6% $17.40 312 trace →
4 GPT 5.5 (high) 52.7% $3.53 40 trace →
5 Claude Sonnet 4.6 49.9% $45.82 705 trace →
6 GLM-5.2 49.8% $21.12 247 trace →
7 Gemini 3.5 Flash 49.7% $6.09 211 trace →
8 Claude Opus 4.7 (xhigh) 46.5% $4.45 107 trace →
9 GPT 5.6 Sol (medium) 45.0% $1.33 17 trace →
10 GPT 5.5 44.9% $1.23 20 trace →
11 Claude Opus 4.7 43.2% $1.26 46 trace →
12 Gemini 3 Flash 43.0% $0.33 117 trace →
13 Gemini 3.6 Flash 38.4% $3.18 147 trace →
14 GPT 5.4 36.7% $0.25 12 trace →
15 GPT 5 mini 36.6% $0.03 23 trace →
16 Gemini 3.1 Pro 34.0% $1.22 91 trace →
17 GPT 5.4 mini 31.8% $0.03 22 trace →
18 Claude Haiku 4.5 28.7% $1.29 165 trace →
19 Claude Opus 4.8 (xhigh) 0.7% $18.78 118 trace →

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