OpenVLM MMT-Bench - Spot The Diff: leaderboard
The "Spot The Diff" subtask of OpenVLM MMT-Bench. OpenCompass OpenVLM evaluation of MMT-Bench: multimodal tasks covering visual recognition, localization, OCR, counting, relation reasoning, visual math, and hallucination detection.
Metric: Score (%). Source: huggingface.co. 207 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Llama 3.2 11B Instruct | 70 | #1126 |
| 2 | Gemma 3 4B | 50 | #1089 |
| 3 | Claude 3 Sonnet | 40 | #775 |
| 4 | Claude 3.5 Sonnet | 35 | #350 |
| 5 | Claude 3.7 Sonnet | 25 | #249 |
| 6 | Claude 3 Haiku | 25 | #803 |
| 7 | MiniCPM-V-2.6 | 20 | #836 |
| 8 | Gemma 3 27B | 20 | #607 |
| 9 | Gemini 1.5 Pro | 20 | #445 |
| 10 | Gemini 1.5 Flash | 20 | #618 |
| 11 | GPT-4.1 Mini | 15 | #355 |
| 12 | Qwen 2 VL 7B | 15 | #750 |
| 13 | InternVL3-14B | 15 | #505 |
| 14 | Pixtral-12B | 15 | #808 |
| 15 | Aquila-VL-2B | 15 | #1238 |
Interactive version: theaggregate.ai/benchmark?slug=openvlm-mmt-bench-spot-the-diff · How It Works · Data refreshed daily, snapshot 2026-10-20.