MEGA-Bench Task - 3D Indoor Scene Text BBox Prediction — leaderboard
Metric: Task Score (%). Source: huggingface.co. 44 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.5 Pro | 28 |
| 2 | Gemini 2.0 Flash (Preview) | 19.4 |
| 3 | GPT-4o | 18.6 |
| 4 | Qwen 2 VL 72B | 16.5 |
| 5 | Gemini 1.5 Pro (002) | 14.7 |
| 6 | InternVL3-78B | 14.6 |
| 7 | InternVL3-38B | 14.5 |
| 8 | Qwen 2 VL 7B | 12 |
| 9 | InternVL2.5-78B | 11.3 |
| 10 | Ovis2-8B | 10.2 |
| 11 | Gemini 1.5 Flash (002) | 9.8 |
| 12 | Claude 3.5 Sonnet (20241022) | 8.1 |
| 13 | Gemma 3 27B (IT) | 6.8 |
| 14 | InternVL3-14B | 6.1 |
| 15 | Claude 3.5 Sonnet (20240620) | 4.7 |
Interactive version: theaggregate.ai/benchmark?slug=mega-bench-task-3d-indoor-scene-text-bbox-prediction · How the rankings work · Data refreshed daily, snapshot 2026-07-22.