WGUNDERWOOD/tex-fmt
An extremely fast LaTeX formatter written in Rust
455
Generated Behavioral Tests
90.8%
Best Score
GPT 5.6 Sol (xhigh)
Results by Model
Hover a point for details · The line marks the Pareto frontier (best score per cost) · Click a point to see model details
Per-Model Results
19 runs| # | Model | Score | Cost | Calls | ||
|---|---|---|---|---|---|---|
| 1 | GPT 5.6 Sol (xhigh) | 90.8% | $7.20 | 49 | trace → | |
| 2 | Claude Opus 4.8 (xhigh) | 89.0% | $26.19 | 174 | trace → | |
| 3 | Claude Opus 4.6 | 80.7% | $19.07 | 374 | trace → | |
| 4 | GLM-5.2 | 77.1% | $25.12 | 238 | trace → | |
| 5 | GPT 5.5 (high) | 76.5% | $3.81 | 48 | trace → | |
| 6 | Gemini 3.5 Flash | 74.3% | $4.23 | 167 | trace → | |
| 7 | Claude Opus 4.7 (xhigh) | 65.1% | $10.97 | 186 | trace → | |
| 8 | Claude Opus 4.7 | 64.0% | $2.87 | 81 | trace → | |
| 9 | GPT 5.5 | 61.8% | $0.78 | 15 | trace → | |
| 10 | Gemini 3.1 Pro | 59.3% | $2.01 | 143 | trace → | |
| 11 | Gemini 3 Flash | 54.5% | $0.25 | 95 | trace → | |
| 12 | GPT 5.4 | 54.3% | $0.50 | 13 | trace → | |
| 13 | Claude Haiku 4.5 | 49.7% | $1.32 | 196 | trace → | |
| 14 | GPT 5.5 (xhigh) | 44.0% | $10.63 | 108 | trace → | |
| 15 | Gemini 3.6 Flash | 41.3% | $3.32 | 172 | trace → | |
| 16 | Claude Sonnet 4.6 | 37.4% | $21.46 | 460 | trace → | |
| 17 | GPT 5.6 Sol (medium) | 37.1% | $0.74 | 11 | trace → | |
| 18 | GPT 5.4 mini | 36.7% | $0.05 | 10 | trace → | |
| 19 | GPT 5 mini | 30.5% | $0.07 | 22 | trace → |
Click a row to replay how that model rebuilt this program, or the model name to open its full run