← Back to leaderboard  ·  Show all task instances

antonmedv/fx

Terminal JSON viewer & processor

20,433 go medium
2,044
Generated Behavioral Tests
93.8%
Best Score
Claude Opus 5 (xhigh)

Results by Model

Hover for details · Click a model to open it · Drag to zoom the x-axis · Shift-drag to pan · The line marks the Pareto frontier (best score per cost)

Per-Model Results

25 runs
# Model Score Cost Calls
1 Claude Opus 5 (xhigh) 93.8% $54.53 295 trace →
2 Muse Spark 1.3 (max) 88.5% $14.05 592 trace →
3 Gemini 3.7 Flash 84.1% $1.50 137 trace →
4 GPT-5.6 Sol (xhigh) 79.9% $5.77 36 trace →
5 Claude Opus 4.8 (xhigh) 78.2% $26.72 182 trace →
6 Claude Opus 4.6 75.7% $20.74 396 trace →
7 GPT 5.5 (high) 74.5% $3.29 50 trace →
8 GLM-5.2 72.9% $19.15 220 trace →
9 Gemini 3.6 Flash 70.6% $4.66 176 trace →
10 GPT-5.6 Sol 69.4% $1.61 22 trace →
11 Muse Spark 1.1 (xhigh) 65.9% $0.25 136 trace →
12 Gemini 3.5 Flash 62.5% $6.59 271 trace →
13 Muse Spark 1.2 (xhigh) 60.6% $0.11 110 trace →
14 GPT 5.5 56.6% $1.21 24 trace →
15 Claude Opus 4.7 49.4% $5.93 145 trace →
16 GPT 5.4 49.3% $0.20 10 trace →
17 Claude Opus 4.7 (xhigh) 47.1% $7.48 135 trace →
18 GPT 5.5 (xhigh) 43.3% $7.63 71 trace →
19 Gemini 3.1 Pro 42.2% $1.37 127 trace →
20 GPT 5.4 mini 40.4% $0.03 7 trace →
21 Claude Haiku 4.5 17.6% $0.73 133 trace →
22 Muse Spark 1.3 (xhigh) 17.4% $1.31 210 trace →
23 Gemini 3 Flash 16.1% $0.19 85 trace →
24 GPT 5 mini 11.6% $0.01 11 trace →
25 Claude Sonnet 4.6 4.3% $25.98 621 trace →

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