OpenVLM MMBench V1.1 EN - Image Topic: leaderboard
OpenCompass OpenVLM MMBench V1.1 English split accuracy for image topic recognition.
Metric: Accuracy (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 284 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Pixtral-12B | 98.9 | #808 |
| 2 | Qwen 2 VL 2B | 98.9 | #1220 |
| 3 | Aquila-VL-2B | 98.9 | #1238 |
| 4 | GPT-5 | 97.8 | #98 |
| 5 | GPT-4.1 Mini | 97.8 | #355 |
| 6 | InternVL3-8B | 97.8 | #620 |
| 7 | GPT-5 Nano | 97.8 | #425 |
| 8 | Gemini 1.5 Flash (002) | 97.8 | #583 |
| 9 | InternVL3-38B | 97.8 | #405 |
| 10 | InternVL3-78B | 97.8 | #352 |
| 11 | InternVL2-8B | 97.8 | #817 |
| 12 | GPT-4.1 (2025-04-14) | 97.8 | #250 |
| 13 | MiniCPM-V-2.6 | 97.8 | #836 |
| 14 | Qwen 2 VL 7B | 97.8 | #750 |
| 15 | InternVL3-14B | 97.8 | #505 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=openvlm-mmbench-v1-1-en-image-topic · How It Works · Data refreshed daily, snapshot 2026-10-20.