MMR-AD - MPDD Detection: leaderboard
Zero-shot industrial anomaly detection on the MPDD images of MMR-AD (2026): given a spatially aligned normal reference image and a test image, the model judges whether the test image shows an anomaly. Accuracy in percent, averaged over product classes; higher is better.
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 July 2028. 11 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | O4 Mini | 70.5 | #368 |
| 2 | InternVL3.5-38B | 70.2 | #1021 |
| 3 | GPT-5 | 68.4 | #178 |
| 4 | InternVL3-8B | 68.3 | #1782 |
| 5 | Qwen 2.5 VL 72B | 67 | #781 |
| 6 | InternVL3-38B | 65.9 | #939 |
| 7 | InternVL3.5-8B | 64.3 | #1604 |
| 8 | Llama 4 Maverick | 64.1 | #1110 |
| 9 | Qwen 2.5 VL 7B | 63.6 | #1862 |
| 10 | GPT-4o | 63.1 | #751 |
| 11 | Gemini 2.5 Pro | 62.1 | #276 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=mmr-ad-mpdd-detection · How It Works · Data refreshed daily, snapshot 2026-11-21.