blacknon/hwatch
A modern alternative to the watch command, records the differences in execution results and can check this differences at after.
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 | 81.1% | $10.71 | 221 |
| 2 |  |
Gemini 3.1 Pro Google | 75.1% | $1.04 | 73 |
| 3 | |
Claude Opus 4.7 Anthropic | 73.5% | $1.23 | 56 |
| 4 | |
Claude Haiku 4.5 Anthropic | 66.0% | $0.59 | 88 |
| 5 | |
Gemini 3 Flash Google | 61.1% | $0.34 | 71 |
| 6 |  |
GPT 5.4 OpenAI | 56.6% | $0.27 | 9 |
| 7 | |
GPT 5.4 mini OpenAI | 24.6% | $0.02 | 9 |
| 8 | |
GPT 5 mini OpenAI | 2.6% | $0.03 | 10 |
| 9 | |
Claude Sonnet 4.6 Anthropic | 1.9% | $14.13 | 378 |
Click row to see model details