anordal/shellharden
The corrective bash syntax highlighter
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 Sonnet 4.6 Anthropic | 81.7% | $74.50 | 840 |
| 2 | |
Claude Opus 4.7 Anthropic | 72.0% | $1.99 | 69 |
| 3 |  |
Gemini 3.1 Pro Google | 71.1% | $1.40 | 143 |
| 4 |  |
GPT 5.4 OpenAI | 63.3% | $0.16 | 9 |
| 5 | |
Gemini 3 Flash Google | 59.6% | $0.24 | 90 |
| 6 | |
Claude Haiku 4.5 Anthropic | 54.4% | $0.96 | 147 |
| 7 | |
GPT 5.4 mini OpenAI | 16.2% | $0.04 | 18 |
| 8 | |
GPT 5 mini OpenAI | 7.2% | $0.03 | 14 |
| 9 | |
Claude Opus 4.6 Anthropic | 0.0% | $2.13 | 256 |
Click row to see model details