SocialReasonBench - Intent Recognition: leaderboard
Metric: Accuracy (%; the 74 intent recognition questions of SocialReasonBench; six-option multiple-choice questions on counterfactual narrative gameplay clips of Detroit: Become Human, labels grounded in the game's script, flowchart and recorded branches; zero-shot, temperature 0 or 0.01; Gemini, Qwen3-Omni, MiniCPM-o and ARC-Hunyuan-Video read native video with audio, GPT-4o, Claude 3.5 Sonnet and GLM-4.6V-Flash 32 evenly sampled keyframes without audio). Source: arxiv.org. Saturation forecast: Around December 2026. 8 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-4o (2024-08-06) | 85.14 |
| 2 | Gemini 3 Pro | 83.78 |
| 3 | Claude 3.5 Sonnet (20240620) | 82.43 |
| 4 | Gemini 3 Flash | 79.73 |
| 5 | Qwen3 Omni 30B A3B Instruct | 72.97 |
Interactive version: theaggregate.ai/benchmark?slug=socialreasonbench-intent-recognition · How It Works · Data refreshed daily, snapshot 2026-09-26.