MEGA-Bench Task - Face Keypoint Detection: leaderboard
Metric: Task Score (%). Source: huggingface.co. 44 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | InternVL2-8B | 84.9 |
| 2 | Aquila-VL-2B | 84.9 |
| 3 | Qwen 2 VL 7B | 75.4 |
| 4 | Gemini 1.5 Flash (002) | 74.6 |
| 5 | Llama 4 Scout Base | 69 |
| 6 | Claude 3.5 Sonnet (20241022) | 68.1 |
| 7 | Gemini 1.5 Pro (002) | 67.6 |
| 8 | Qwen 2 VL 72B | 67 |
| 9 | GPT-4o Mini | 65 |
| 10 | Pixtral-12B | 64.9 |
| 11 | InternVL3-38B | 63.3 |
| 12 | Gemini 2.5 Pro (Preview 03-25) | 61.7 |
| 13 | Gemma 3 27B (IT) | 60.4 |
| 14 | Claude 3.5 Sonnet (20240620) | 59.9 |
| 15 | GPT-4o | 59.8 |
Interactive version: theaggregate.ai/benchmark?slug=mega-bench-task-face-keypoint-detection · How It Works · Data refreshed daily, snapshot 2026-09-05.