MEGA-Bench Task - Table-to-LaTeX Complex: leaderboard
Metric: Task Score (%). Source: huggingface.co. 44 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.5 Pro (Preview 03-25) | 84.4 |
| 2 | Gemini 1.5 Flash (002) | 76.7 |
| 3 | Qwen 2 VL 72B | 75.6 |
| 4 | InternVL3-78B | 73.3 |
| 5 | Claude 3.5 Sonnet (20240620) | 72.2 |
| 6 | Gemini 1.5 Pro (002) | 71.1 |
| 7 | InternVL3-38B | 70 |
| 8 | Claude 3.5 Sonnet (20241022) | 68.9 |
| 9 | GPT-4o | 67.8 |
| 10 | InternVL2.5-78B | 66.7 |
| 11 | InternVL3-8B | 64.4 |
| 12 | Gemma 3 27B (IT) | 63.3 |
| 13 | InternVL2-8B | 62.2 |
| 14 | GPT-4o Mini | 56.7 |
| 15 | Qwen 2 VL 7B | 53.3 |
Interactive version: theaggregate.ai/benchmark?slug=mega-bench-task-table-to-latex-complex · How It Works · Data refreshed daily, snapshot 2026-09-05.