OpenVLM CCBench - Traditional Shows: leaderboard
Metric: Accuracy (%). Source: huggingface.co. 228 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | InternVL3-8B | 90.9 |
| 2 | InternVL3-78B | 87.9 |
| 3 | InternVL3-14B | 87.9 |
| 4 | InternVL2-8B | 78.8 |
| 5 | Qwen 2 VL 72B | 69.7 |
| 6 | GPT-4.1 (2025-04-14) | 68.2 |
| 7 | GPT-4.1 Mini | 65.2 |
| 8 | Qwen 2 VL 7B | 63.6 |
| 9 | Moonshot v1 8K | 63.6 |
| 10 | Grok 2 (1212) | 62.1 |
| 11 | Llama 3.2 90B Vision Instruct | 56.1 |
| 12 | Gemini 1.5 Flash | 54.5 |
| 13 | Gemini 1.0 Pro | 54.5 |
| 14 | Claude 3.5 Sonnet (20241022) | 53 |
| 15 | Qwen 2 VL 2B | 53 |
Interactive version: theaggregate.ai/benchmark?slug=openvlm-ccbench-traditional-shows · How It Works · Data refreshed daily, snapshot 2026-09-19.