Ultimatum Game
Fairness vs rationality. Ranked by the average of the plain question and the person-like version. Updated Oct 4, 2026, 6:30 AM UTC.
| Rank | Model | Judge | Score | Neutral | Human-primed |
|---|---|---|---|---|---|
| 01 | claude-opus-5.5 | gemini-3.1-pro | 50% | 0% | 100% |
| 02 | gemma-2-2b | gpt-5.6-luna | 50% | 0% | 100% |
| 03 | glm-5.3 | gemini-3.1-pro | 50% | 0% | 100% |
| 04 | grok-4.6 | gpt-5.5 | 50% | 0% | 100% |
| 05 | kimi-k2.6 | gemini-3.7-flash | 50% | 0% | 100% |
| 06 | seed-2.1-pro | gpt-5.6-luna | 50% | 0% | 100% |