WGUNDERWOOD/tex-fmt
An extremely fast LaTeX formatter written in Rust
Results by Model
| # | Model | Score Percentage of hidden behavioral tests passed. | Cost Total API cost in USD for this task instance. | Calls Number of LLM API calls for this task instance. | |
|---|---|---|---|---|---|
| 1 | Claude Opus 4.6 Anthropic | 80.7% | $19.07 | 374 | |
| 2 | Claude Opus 4.7 Anthropic | 64.0% | $2.87 | 81 | |
| 3 | Gemini 3.1 Pro Google | 59.3% | $2.01 | 143 | |
| 4 | Gemini 3 Flash Google | 54.5% | $0.25 | 95 | |
| 5 | GPT 5.4 OpenAI | 54.3% | $0.50 | 13 | |
| 6 | Claude Haiku 4.5 Anthropic | 49.7% | $1.32 | 196 | |
| 7 | Claude Sonnet 4.6 Anthropic | 37.4% | $21.46 | 460 | |
| 8 | GPT 5.4 mini OpenAI | 36.7% | $0.05 | 10 | |
| 9 | GPT 5 mini OpenAI | 30.5% | $0.07 | 22 |
Click row to see model details