HaliteChallenge/Halite
@twosigma's first artificial intelligence programming challenge
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 | 80.4% | $15.08 | 222 |
| 2 |  |
Gemini 3.1 Pro Google | 75.6% | $1.56 | 73 |
| 3 | |
Claude Opus 4.7 Anthropic | 71.3% | $1.00 | 44 |
| 4 | |
Gemini 3 Flash Google | 47.3% | $0.52 | 71 |
| 5 | |
Claude Haiku 4.5 Anthropic | 42.9% | $0.45 | 106 |
| 6 |  |
GPT 5.4 OpenAI | 18.5% | $0.20 | 6 |
| 7 | |
Claude Sonnet 4.6 Anthropic | 14.2% | $38.69 | 494 |
| 8 | |
GPT 5 mini OpenAI | 1.5% | $0.04 | 22 |
| 9 | |
GPT 5.4 mini OpenAI | 0.0% | $0.02 | 9 |
Click row to see model details