MEGA-Bench Task - COCO Object Detection By Query Property — leaderboard
Metric: Task Score (%). Source: huggingface.co. 44 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.5 Pro | 78.7 |
| 2 | InternVL2.5-78B | 77.7 |
| 3 | Qwen 2 VL 72B | 74 |
| 4 | InternVL3-38B | 71.4 |
| 5 | Gemini 2.0 Flash (Preview) | 67 |
| 6 | Gemma 3 27B (IT) | 65.1 |
| 7 | InternVL3-78B | 64.7 |
| 8 | InternVL3-14B | 62.9 |
| 9 | Qwen 2 VL 7B | 62.5 |
| 10 | Llama 4 Scout Base | 62.2 |
| 11 | InternVL3-8B | 61.9 |
| 12 | Claude 3.5 Sonnet (20241022) | 60.6 |
| 13 | Gemini 1.5 Pro (002) | 59.5 |
| 14 | GPT-4o | 58.1 |
| 15 | Gemini 1.5 Flash (002) | 58 |
Interactive version: theaggregate.ai/benchmark?slug=mega-bench-task-coco-object-detection-by-query-property · How the rankings work · Data refreshed daily, snapshot 2026-07-22.