MEGA-Bench Task - Visual Prediction Rater Panoptic Segmentation: leaderboard
Metric: Task Score (%). Source: huggingface.co. 44 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.5 Pro (Preview 03-25) | 88.1 |
| 2 | Claude 3.5 Sonnet (20241022) | 81 |
| 3 | Claude 3.5 Sonnet (20240620) | 59.5 |
| 4 | Gemini 1.5 Pro (002) | 50 |
| 5 | GPT-4o | 47.6 |
| 6 | InternVL3-78B | 42.9 |
| 7 | Gemma 3 27B (IT) | 38.1 |
| 8 | Gemini 1.5 Flash (002) | 26.2 |
| 9 | Llama 4 Scout Base | 21.4 |
| 10 | InternVL3-38B | 19 |
| 11 | Gemma 3 12B (IT) | 14.3 |
| 12 | Gemma 3 4B (IT) | 9.5 |
| 13 | Qwen 2 VL 2B | 9.5 |
| 14 | Pixtral-12B | 9.5 |
| 15 | InternVL3-8B | 7.1 |
Interactive version: theaggregate.ai/benchmark?slug=mega-bench-task-visual-prediction-rater-panoptic-segmentation · How It Works · Data refreshed daily, snapshot 2026-09-05.