WildRoadBench - Road Crack: leaderboard
Metric: F1 (%) of predicted boxes at IoU 0.5 for the road crack scene class (561 ground-truth boxes), VLM Track, zero-shot box grounding on the full 1,061-image aerial set (1,394 ground-truth boxes); higher is better. Source: arxiv.org. Saturation forecast: Around March 2028. 25 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3 Pro | 33.3 |
| 2 | Qwen 3.6 Plus | 25.4 |
| 3 | Gemini 3.1 Flash Lite | 16.2 |
| 4 | Qwen 2.5 VL 7B Instruct | 15.3 |
| 5 | Claude Sonnet 4.6 | 11.3 |
| 6 | GPT-5.4 (2026-03-05) | 10.8 |
| 7 | Qwen 3 VL 8B Instruct | 6.2 |
| 8 | Kimi K2.5 | 4 |
| 9 | Claude Opus 4.6 | 2.1 |
| 10 | Gemini 2.5 Flash | 1.9 |
| 11 | Qwen 3 VL 8B (Thinking) | 1.9 |
| 12 | InternVL3-8B | 0.8 |
| 13 | GPT-4o (2024-05-13) | 0.7 |
| 14 | InternVL3-14B | 0.6 |
| 15 | InternVL2-8B | 0 |
Interactive version: theaggregate.ai/benchmark?slug=wildroadbench-road-crack · How It Works · Data refreshed daily, snapshot 2026-10-07.