PuMVR - Roman: leaderboard
Metric: Zero-shot Accuracy (%; Roman script). Source: arxiv.org. Saturation forecast: Estimated already saturated. 10 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.5 Flash | 92.27 |
| 2 | Grok 4.1 Fast (Reasoning) | 90.13 |
| 3 | GPT-4o | 86.93 |
| 4 | Claude Sonnet 4 | 85.83 |
| 5 | Qwen 2 VL 72B Instruct | 80.25 |
| 6 | Qwen 2 VL 7B Instruct | 65.07 |
| 7 | Llama 3.2 11B Instruct | 60.27 |
Interactive version: theaggregate.ai/benchmark?slug=pumvr-roman · How It Works · Data refreshed daily, snapshot 2026-09-25.