OpenVLM Q-Bench - Yes-or-No (Others): leaderboard
Metric: Accuracy (%). Source: huggingface.co. 160 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | InternVL3-14B | 79.4 |
| 2 | Qwen 2 VL 72B | 78.9 |
| 3 | GPT-4.1 (2025-04-14) | 78.3 |
| 4 | InternVL3-8B | 78.3 |
| 5 | InternVL3-78B | 78.3 |
| 6 | Qwen 2 VL 7B | 77.7 |
| 7 | Moonshot v1 8K | 77.1 |
| 8 | GPT-4.1 Mini | 76.6 |
| 9 | Qwen 2 VL 2B | 76.6 |
| 10 | Pixtral-12B | 76 |
| 11 | InternVL2-8B | 74.3 |
| 12 | Gemini 1.5 Flash | 72.6 |
| 13 | Gemma 3 27B | 72 |
| 14 | Gemma 3 12B | 72 |
| 15 | GPT-4.1 Nano | 71.4 |
Interactive version: theaggregate.ai/benchmark?slug=openvlm-q-bench-yes-or-no-others · How It Works · Data refreshed daily, snapshot 2026-09-19.