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) 84.6% $3.74 27 trace →
2 Claude Opus 4.8 (xhigh) 78.3% $25.11 174 trace →
3 GPT 5.5 (xhigh) 77.9% $5.02 41 trace →
4 GLM-5.2 76.3% $25.23 186 trace →
5 Claude Sonnet 4.6 72.1% $19.74 435 trace →
6 Gemini 3.5 Flash 71.3% $3.69 91 trace →
7 GPT 5.5 (high) 70.6% $2.55 28 trace →
8 GPT 5.6 Sol (medium) 70.6% $0.72 13 trace →
9 GPT 5.5 68.0% $1.16 13 trace →
10 Claude Opus 4.7 (xhigh) 61.1% $4.10 75 trace →
11 Gemini 3.6 Flash 54.6% $3.62 111 trace →
12 GPT 5.4 52.6% $0.28 8 trace →
13 Claude Opus 4.7 46.0% $1.40 39 trace →
14 Gemini 3.1 Pro 33.9% $0.96 43 trace →
15 Gemini 3 Flash 26.3% $0.29 55 trace →
16 GPT 5.4 mini 15.8% $0.03 7 trace →
17 GPT 5 mini 12.8% $0.02 8 trace →
18 Claude Opus 4.6 0.0% $2.89 187 trace →
19 Claude Haiku 4.5 0.0% $0.93 99 trace →

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