GroupToM-Bench - Group Tension: leaderboard
GroupToM-Bench (2026) tests group-level theory of mind in multimodal models: 240 expert-seeded multi-agent scenarios, each a scene image with dialogue, are probed along a seven-level audit from individual mental states to collective outcomes. This board is level 4 group tension detection. Scored by strict exact-match accuracy in percent on multi-select multiple choice (random 6.7); higher is better.
Metric: strict exact-match accuracy (%) over multi-select options (random 6.7) on Level 4 group tension detection (false consensus and latent subgroup antagonism), multiple choice; 240 multimodal multi-agent scenarios (scene image plus dialogue); higher is better. Source: arxiv.org. Saturation forecast: Around 2029. 11 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Pro | 53.1 | #137 |
| 2 | GPT-5 | 50.5 | #167 |
| 3 | GPT-4o | 50.3 | #741 |
| 4 | Claude Haiku 4.5 | 50.2 | #568 |
| 5 | GPT-5 Mini | 49.4 | #348 |
| 6 | Llama 3.2 11B | 39.8 | #6383 |
| 7 | GPT-5 Nano | 38 | #889 |
| 8 | Qwen 2.5 VL 7B Instruct | 36.4 | #1833 |
| 9 | Qwen 2 VL 7B | 26.2 | #2261 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=grouptom-bench-group-tension · How It Works · Data refreshed daily, snapshot 2026-11-13.