alecthomas/chroma

A general purpose syntax highlighter in pure Go

4,910 go
503
Generated Behavioral Tests
46.5%
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) 46.5% $40.55 212 trace →
2 GPT 5.5 (high) 41.7% $4.27 54 trace →
3 GPT 5.5 27.2% $1.51 22 trace →
4 Gemini 3.6 Flash 26.2% $6.32 184 trace →
5 Claude Sonnet 4.6 16.1% $37.36 565 trace →
6 GLM-5.2 14.5% $39.91 299 trace →
7 Gemini 3.7 Flash 13.5% $3.41 185 trace →
8 GPT 5.5 (xhigh) 13.1% $9.46 80 trace →
9 Claude Opus 4.8 (xhigh) 11.1% $9.48 84 trace →
10 GPT 5.4 9.3% $0.33 9 trace →
11 Claude Opus 4.7 (xhigh) 8.7% $5.07 126 trace →
12 Gemini 3 Flash 7.6% $0.24 64 trace →
13 Claude Haiku 4.5 6.2% $0.65 114 trace →
14 Claude Opus 4.7 4.6% $0.48 20 trace →
15 Gemini 3.1 Pro 4.4% $1.90 141 trace →
16 GPT 5.4 mini 2.2% $0.03 8 trace →
17 GPT-5.6 Sol 1.2% $1.73 15 trace →
18 GPT 5 mini 1.2% $0.02 16 trace →
19 GPT-5.6 Sol (xhigh) 0.6% $7.57 42 trace →
20 Claude Opus 4.6 0.6% $16.38 297 trace →
21 Gemini 3.5 Flash 0.0% $14.88 283 trace →

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