MMR-AD - MPDD Detection: leaderboard
Metric: Anomaly detection accuracy (%): share of MPDD test images (from MMR-AD) correctly judged normal or anomalous, zero-shot, the model sees a spatially aligned normal reference image and the test image with the MMR-AD instruction template; dataset-level average over the subdataset's product classes; higher is better. Source: arxiv.org. Saturation forecast: Around 2029. 13 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | O4 Mini | 70.5 |
| 2 | GPT-5 | 68.4 |
| 3 | InternVL3-8B | 68.3 |
| 4 | InternVL3-38B | 65.9 |
| 5 | Llama 4 Maverick | 64.1 |
| 6 | Qwen 2.5 VL 7B | 63.6 |
| 7 | GPT-4o | 63.1 |
| 8 | Gemini 2.5 Pro | 62.1 |
Interactive version: theaggregate.ai/benchmark?slug=mmr-ad-mpdd-detection · How It Works · Data refreshed daily, snapshot 2026-10-07.