OpenVLM MMT-Bench - Writing Poetry From Image — leaderboard
Metric: Score (%). Source: huggingface.co. 207 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-4.1 (2025-04-14) | 85 |
| 2 | Llama 3.2 90B Vision Instruct | 85 |
| 3 | InternVL3-38B | 80 |
| 4 | InternVL3-78B | 75 |
| 5 | Qwen 2 VL 72B | 75 |
| 6 | Claude 3.5 Sonnet | 75 |
| 7 | Grok 2 (1212) | 70 |
| 8 | InternVL3-8B | 65 |
| 9 | InternVL3-14B | 65 |
| 10 | GPT-4.1 Mini | 60 |
| 11 | Pixtral-12B | 60 |
| 12 | Llama 3.2 11B Instruct | 60 |
| 13 | Gemini 1.5 Flash | 60 |
| 14 | Gemma 3 27B | 55 |
| 15 | Claude 3 Haiku | 55 |
Interactive version: theaggregate.ai/benchmark?slug=openvlm-mmt-bench-writing-poetry-from-image · How the rankings work · Data refreshed daily, snapshot 2026-07-22.