OpenVLM MMT-Bench - Industrial Produce Anomaly Detection: leaderboard
Metric: Score (%). Source: huggingface.co. 207 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen 2 VL 72B | 90 |
| 2 | Gemini 1.5 Pro | 90 |
| 3 | Qwen 2 VL 2B | 85 |
| 4 | Aquila-VL-2B | 80 |
| 5 | Pixtral-12B | 80 |
| 6 | Claude 3.5 Sonnet | 80 |
| 7 | Claude 3 Haiku | 80 |
| 8 | InternVL3-38B | 75 |
| 9 | Llama 3.2 90B Vision Instruct | 75 |
| 10 | InternVL3-78B | 70 |
| 11 | Claude 3.7 Sonnet | 65 |
| 12 | GPT-4.1 (2025-04-14) | 65 |
| 13 | Qwen 2 VL 7B | 60 |
| 14 | Gemini 1.5 Flash | 60 |
| 15 | Llama 3.2 11B Instruct | 55 |
Interactive version: theaggregate.ai/benchmark?slug=openvlm-mmt-bench-industrial-produce-anomaly-detection · How It Works · Data refreshed daily, snapshot 2026-09-05.