MARINER - Vessel Detection (Civilian): leaderboard
Metric: Average precision of MARINER vessel detection by multimodal LLMs (COCO-style average precision over IoU 0.5 to 0.95, percent; civilian vessel categories); higher is better. Source: arxiv.org. Saturation forecast: Around January 2027. 15 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.5 Pro | 48.47 |
| 2 | InternVL3-38B | 33.07 |
| 3 | InternVL3-78B | 19.32 |
| 4 | MiniCPM-V-2.6 | 12.87 |
| 5 | InternVL3-8B | 11.87 |
| 6 | Gemini 2.5 Flash | 10.14 |
| 7 | GPT-4.1 | 9.14 |
| 8 | Qwen 2.5 VL 32B Instruct | 6 |
| 9 | Qwen 2.5 VL 72B Instruct | 5.64 |
| 10 | Qwen 2.5 VL 7B Instruct | 5.4 |
| 11 | GPT-4o | 3.12 |
| 12 | InternVL2-8B | 1.81 |
Interactive version: theaggregate.ai/benchmark?slug=mariner-vessel-detection-civilian · How It Works · Data refreshed daily, snapshot 2026-10-07.