MARINER - Vessel Detection (Military): leaderboard
Vessel detection by multimodal language models on MARINER (2026), a maritime vision benchmark of shore, shipboard, drone and open-source images. COCO-style average precision in percent over military vessel categories; higher is better.
Metric: Average precision of MARINER vessel detection by multimodal LLMs (COCO-style average precision over IoU 0.5 to 0.95, percent; military vessel categories); higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 13 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 2.5 Pro | 69.04 | #280 |
| 2 | Qwen 2.5 VL 72B Instruct | 37.29 | #849 |
| 3 | Qwen 2.5 VL 32B Instruct | 26.91 | #1065 |
| 4 | Qwen 2.5 VL 7B Instruct | 26.1 | #1919 |
| 5 | Gemini 2.5 Flash | 22.91 | #466 |
| 6 | InternVL3-38B | 21.77 | #948 |
| 7 | Qwen 2.5 VL 3B Instruct | 8.35 | #2720 |
| 8 | InternVL3-78B | 7.91 | #765 |
| 9 | LLaVA-OneVision-7B | 7.59 | #3002 |
| 10 | LLaVA-OneVision-72B | 7.35 | #1782 |
| 11 | MiniCPM-V-2.6 | 5.93 | #3311 |
| 12 | InternVL3-8B | 2.37 | #1795 |
| 13 | InternVL2-8B | 0.43 | #3268 |
No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=mariner-vessel-detection-military · How It Works · Data refreshed daily, snapshot 2026-11-22.