MEGA-Bench Task - Newspaper Page Parse And Count — leaderboard
Metric: Task Score (%). Source: huggingface.co. 44 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-4o | 64.4 |
| 2 | Gemini 2.5 Pro | 62.2 |
| 3 | InternVL2.5-78B | 60 |
| 4 | InternVL2-8B | 53.3 |
| 5 | Qwen 2 VL 72B | 53.3 |
| 6 | Claude 3.5 Sonnet (20240620) | 53.3 |
| 7 | InternVL3-78B | 51.1 |
| 8 | InternVL3-38B | 48.9 |
| 9 | Qwen 2 VL 7B | 46.7 |
| 10 | InternVL3-14B | 46.7 |
| 11 | Gemini 1.5 Pro (002) | 46.7 |
| 12 | Llama 4 Scout Base | 46.7 |
| 13 | Claude 3.5 Sonnet (20241022) | 44.4 |
| 14 | InternVL3-8B | 42.2 |
| 15 | Gemini 2.0 Flash (Preview) | 40 |
Interactive version: theaggregate.ai/benchmark?slug=mega-bench-task-newspaper-page-parse-and-count · How the rankings work · Data refreshed daily, snapshot 2026-07-22.