OpenVLM MMT-Bench - Spot The Diff: leaderboard
Metric: Score (%). Source: huggingface.co. 207 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Llama 3.2 11B Instruct | 70 |
| 2 | Gemma 3 4B | 50 |
| 3 | Claude 3 Sonnet | 40 |
| 4 | Claude 3.5 Sonnet | 35 |
| 5 | Claude 3.7 Sonnet | 25 |
| 6 | Claude 3 Haiku | 25 |
| 7 | Gemma 3 27B | 20 |
| 8 | Gemini 1.5 Pro | 20 |
| 9 | Gemini 1.5 Flash | 20 |
| 10 | GPT-4.1 Mini | 15 |
| 11 | Qwen 2 VL 7B | 15 |
| 12 | Aquila-VL-2B | 15 |
| 13 | Pixtral-12B | 15 |
| 14 | Gemma 3 12B | 15 |
| 15 | Llama 3.2 90B Vision Instruct | 15 |
Interactive version: theaggregate.ai/benchmark?slug=openvlm-mmt-bench-spot-the-diff · How It Works · Data refreshed daily, snapshot 2026-09-05.