A3Bench - Basic Questions: leaderboard
Zero-shot multiple-choice video QA accuracy (%) on single-attribute basic questions in A3Bench (arXiv:2611.24762).
Metric: Basic Questions Accuracy (%). Source: arxiv.org. Saturation forecast: Around February 2028. 14 models tracked. Updated 20 Nov 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | InternVL2.5-8B | 52.61 | #2370 |
| 2 | Qwen 3 VL 8B Instruct | 51.07 | #1035 |
| 3 | Ovis2.5-9B | 50.62 | #1189 |
| 4 | InternVL3-9B | 50.05 | #1731 |
| 5 | VideoLLaMA3-7B | 49.65 | #2000 |
| 6 | Gemma 4 26B A4B (IT) | 49.07 | #561 |
| 7 | InternVL3.5-8B | 48.34 | #1599 |
| 8 | InternVideo2.5-8B | 48.07 | |
| 9 | GPT-5.1 | 45.63 | #242 |
| 10 | Qwen 3 VL 8B (Thinking) | 44.59 | #920 |
| 11 | Qwen 2.5 VL 7B Instruct | 41.87 | #1904 |
| 12 | ViCLIP | 28.11 | |
| 13 | X-CLIP | 23.67 | |
| 14 | LLaVA-Video-7B-Qwen2 | 19.58 | #1977 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=a3bench-basic-questions · How It Works · Data refreshed daily, snapshot 2026-11-20.