MEGA-Bench - Input Format: User Interface Screenshots: leaderboard
Metric: Average Score (%). Source: huggingface.co. 44 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.5 Pro (Preview 03-25) | 68.9 |
| 2 | GPT-4o | 60.8 |
| 3 | Claude 3.5 Sonnet (20241022) | 60.5 |
| 4 | Claude 3.5 Sonnet (20240620) | 56.9 |
| 5 | Qwen 2 VL 72B | 52.6 |
| 6 | InternVL2.5-78B | 49.1 |
| 7 | GPT-4o Mini | 47.2 |
| 8 | InternVL3-78B | 46.4 |
| 9 | Gemini 1.5 Pro (002) | 46.3 |
| 10 | InternVL3-38B | 42.5 |
| 11 | Gemini 1.5 Flash (002) | 39.7 |
| 12 | Gemma 3 27B (IT) | 39.6 |
| 13 | InternVL3-8B | 35.2 |
| 14 | Qwen 2 VL 7B | 35.2 |
| 15 | Gemma 3 12B (IT) | 33.7 |
Interactive version: theaggregate.ai/benchmark?slug=mega-bench-input-format-user-interface-screenshots · How It Works · Data refreshed daily, snapshot 2026-09-05.