MEGA-Bench Task - GUI Act Web Multi: leaderboard
Metric: Task Score (%). Source: huggingface.co. 44 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 1.5 Pro (002) | 47 |
| 2 | InternVL2.5-78B | 46.4 |
| 3 | Claude 3.5 Sonnet (20240620) | 46.4 |
| 4 | Gemini 2.5 Pro (Preview 03-25) | 45.7 |
| 5 | Qwen 2 VL 72B | 44.1 |
| 6 | GPT-4o | 43.8 |
| 7 | Claude 3.5 Sonnet (20241022) | 42.9 |
| 8 | InternVL3-38B | 42.9 |
| 9 | Qwen 2 VL 2B | 42.9 |
| 10 | InternVL2-8B | 39.3 |
| 11 | InternVL3-8B | 37.4 |
| 12 | Qwen 2 VL 7B | 35.9 |
| 13 | Gemma 3 27B (IT) | 35.7 |
| 14 | InternVL3-78B | 25 |
| 15 | Gemini 1.5 Flash (002) | 22.3 |
Interactive version: theaggregate.ai/benchmark?slug=mega-bench-task-gui-act-web-multi · How It Works · Data refreshed daily, snapshot 2026-09-05.