UAVQA-Bench - Fine-Grained Attribute Perception: leaderboard
Metric: Accuracy (%; attribute, function and safe-landing recognition; direct answering over UAVQA-Bench: 1,500 human-annotated questions on images from 13 public UAV datasets, 16 tasks in six capability dimensions, multiple choice plus box grounding counted correct at IoU 0.5). Source: arxiv.org. Saturation forecast: Around December 2026. 14 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3 Pro | 68.5 |
| 2 | Gemini 3 Flash | 60.5 |
| 3 | Qwen 3 VL 32B Instruct | 58 |
| 4 | Qwen 3.5 9B | 57.5 |
| 5 | Qwen 3 VL 32B (Thinking) | 57 |
| 6 | Qwen 3 VL 30B A3B (Thinking) | 53.5 |
| 7 | Qwen 3 VL 30B A3B Instruct | 51.5 |
| 8 | Qwen 3 VL 8B Instruct | 46.5 |
| 9 | Qwen 3 VL 8B (Thinking) | 44.5 |
| 10 | InternVL3.5-8B | 36.5 |
Interactive version: theaggregate.ai/benchmark?slug=uavqa-bench-fine-grained-attribute-perception · How It Works · Data refreshed daily, snapshot 2026-09-29.