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) 10.4% $7.13 45 trace →
2 Gemini 3.1 Pro 9.4% $1.82 100 trace →
3 Gemini 3.6 Flash 8.1% $4.11 120 trace →
4 Claude Opus 4.8 (xhigh) 7.6% $8.71 87 trace →
5 GLM-5.2 7.4% $13.98 178 trace →
6 GPT 5.5 (high) 7.2% $5.07 51 trace →
7 Claude Sonnet 4.6 5.0% $32.36 562 trace →
8 Claude Opus 4.7 4.5% $6.17 153 trace →
9 Claude Opus 4.6 4.3% $14.33 298 trace →
10 GPT 5.5 4.3% $0.80 11 trace →
11 GPT 5.6 Sol (medium) 3.6% $0.71 11 trace →
12 GPT 5.4 3.2% $0.32 9 trace →
13 GPT 5.5 (xhigh) 3.2% $7.86 62 trace →
14 GPT 5.4 mini 2.4% $0.04 8 trace →
15 Gemini 3.5 Flash 1.5% $0.83 63 trace →
16 Claude Haiku 4.5 1.4% $1.04 97 trace →
17 Claude Opus 4.7 (xhigh) 1.1% $7.41 151 trace →
18 GPT 5 mini 0.9% $0.02 11 trace →
19 Gemini 3 Flash 0.0% $0.34 71 trace →

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