MARINER - VQA Spatial Relationship: leaderboard
Maritime visual question answering on MARINER (2026), with multiple-choice questions built from verified image metadata and audited by maritime annotators. Accuracy in percent on spatial relationship questions; higher is better.
Metric: Accuracy (%) on MARINER maritime visual question answering (spatial relationship questions; multiple-choice questions generated from verified image metadata and audited by maritime annotators); higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 16 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-4o | 35.26 | #751 |
| 2 | Qwen 2.5 VL 72B Instruct | 35.26 | #841 |
| 3 | Qwen 2.5 VL 32B Instruct | 35.26 | #1062 |
| 4 | Gemini 2.5 Flash | 33.53 | #463 |
| 5 | InternVL3-38B | 33.53 | #939 |
| 6 | GPT-4.1 | 32.37 | #505 |
| 7 | Gemini 2.5 Pro | 31.79 | #276 |
| 8 | InternVL2-8B | 31.79 | #3227 |
| 9 | InternVL3-78B | 30.64 | #759 |
| 10 | Qwen 2.5 VL 3B Instruct | 30.64 | #2713 |
| 11 | InternVL3-8B | 29.48 | #1782 |
| 12 | Qwen 2.5 VL 7B Instruct | 27.75 | #1902 |
| 13 | LLaVA-OneVision-72B | 27.75 | #1771 |
| 14 | MiniCPM-V-2.6 | 26.01 | #3267 |
| 15 | LLaVA-OneVision-7B | 24.86 | #2981 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=mariner-vqa-spatial-relationship · How It Works · Data refreshed daily, snapshot 2026-11-21.