ggreer/the_silver_searcher

A code-searching tool similar to ack, but faster.

27,080 c medium
1,004
Generated Behavioral Tests
97.0%
Best Score
Claude Opus 5 (xhigh)

Hover a point for details · The line marks the Pareto frontier (best score per cost) · Click a point to see model details

21 runs
# Model Score Cost Calls
1 Claude Opus 5 (xhigh) 97.0% $52.77 247 trace →
2 GPT-5.6 Sol (xhigh) 88.8% $4.01 29 trace →
3 GPT 5.5 (xhigh) 85.8% $9.59 98 trace →
4 Claude Opus 4.8 (xhigh) 79.1% $11.74 95 trace →
5 GLM-5.2 78.8% $16.15 218 trace →
6 Claude Opus 4.7 (xhigh) 78.4% $6.00 113 trace →
7 GPT 5.5 (high) 76.7% $4.66 52 trace →
8 Gemini 3.7 Flash 75.2% $1.14 117 trace →
9 GPT-5.6 Sol 71.5% $1.03 15 trace →
10 Claude Opus 4.6 70.2% $7.38 195 trace →
11 Gemini 3.5 Flash 63.1% $7.35 218 trace →
12 GPT 5.5 59.7% $1.54 20 trace →
13 Claude Sonnet 4.6 59.5% $13.76 443 trace →
14 Gemini 3.1 Pro 55.7% $1.38 103 trace →
15 GPT 5.4 51.8% $0.20 6 trace →
16 Claude Haiku 4.5 38.9% $0.88 144 trace →
17 Gemini 3 Flash 26.6% $0.33 91 trace →
18 GPT 5 mini 10.6% $0.01 9 trace →
19 GPT 5.4 mini 9.6% $0.39 323 trace →
20 Gemini 3.6 Flash 9.5% $4.13 164 trace →
21 Claude Opus 4.7 1.5% $3.72 107 trace →

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