OpenVLM MMT-Bench - Social Relation Recognition: leaderboard
The "Social Relation Recognition" 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. Saturation forecast: Estimated already saturated. 207 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Grok 2 (1212) | 75 | #496 |
| 2 | GPT-4.1 Mini | 65 | #355 |
| 3 | InternVL3-14B | 65 | #505 |
| 4 | Gemma 3 27B | 65 | #607 |
| 5 | GPT-4.1 (2025-04-14) | 60 | #250 |
| 6 | Gemini 1.5 Pro | 60 | #445 |
| 7 | Aquila-VL-2B | 60 | #1238 |
| 8 | Gemini 1.5 Flash | 60 | #618 |
| 9 | Gemma 3 4B | 55 | #1089 |
| 10 | Llama 3.2 11B Instruct | 55 | #1126 |
| 11 | Claude 3.7 Sonnet | 50 | #249 |
| 12 | Pixtral-12B | 50 | #808 |
| 13 | Qwen 2 VL 2B | 50 | #1220 |
| 14 | InternVL3-8B | 45 | #620 |
| 15 | Claude 3.5 Sonnet | 45 | #350 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=openvlm-mmt-bench-social-relation-recognition · How It Works · Data refreshed daily, snapshot 2026-10-20.