MEGA-Bench Task - Weather Info Parsing: leaderboard
Metric: Task Score (%). Source: huggingface.co. 44 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen 2 VL 7B | 97.2 |
| 2 | Qwen 2 VL 72B | 96 |
| 3 | Gemini 2.5 Pro (Preview 03-25) | 94.8 |
| 4 | GPT-4o | 92.9 |
| 5 | InternVL3-78B | 92.9 |
| 6 | InternVL3-38B | 92.9 |
| 7 | InternVL3-8B | 92.1 |
| 8 | InternVL2.5-78B | 92.1 |
| 9 | Claude 3.5 Sonnet (20241022) | 91.7 |
| 10 | Claude 3.5 Sonnet (20240620) | 90.9 |
| 11 | GPT-4o Mini | 87.7 |
| 12 | Llama 4 Scout Base | 83.7 |
| 13 | Gemini 1.5 Pro (002) | 81.7 |
| 14 | Gemini 1.5 Flash (002) | 75.4 |
| 15 | Qwen 2 VL 2B | 73.4 |
Interactive version: theaggregate.ai/benchmark?slug=mega-bench-task-weather-info-parsing · How It Works · Data refreshed daily, snapshot 2026-09-05.