BehaviorBench: leaderboard
Foundation models have been increasingly applied to behavioral science domains such as psychology, sociology, and economics. While these models show promise in individual tasks such as survey response prediction and human-subject experiment simulation, there remains no systematic understanding of how well they perform across diverse behavioral science tasks, contexts, and populations. We introduce
Metric: Game Behav. Sim. (W ↓) (self-reported). Source: benchmarklist.com. Saturation forecast: Not forecast. 20 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | DeepSeek V3.2 | 16.9 | #205 |
| 2 | Gemini 3.1 Pro (Preview) | 21.3 | #58 |
| 3 | Llama 3.3 70B | 24.2 | #579 |
| 4 | Claude Haiku 4.5 | 24.5 | #280 |
| 5 | Claude Sonnet 4.6 | 24.6 | #90 |
| 6 | Claude Opus 4.6 | 26 | #63 |
| 7 | Qwen 3 4B | 26.9 | #829 |
| 8 | GPT-4.1 | 27.5 | #248 |
| 9 | GPT-5.4 Mini | 29 | #212 |
| 10 | Gemini 3.1 Flash Lite (Preview) | 31.3 | #194 |
| 11 | GPT-5.4 | 31.4 | #79 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=behaviorbench · How It Works · Data refreshed daily, snapshot 2026-10-20.