MEGA-Bench Task - SciBench With Solution Open Ended — leaderboard
Metric: Task Score (%). Source: huggingface.co. 44 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.5 Pro | 72.6 |
| 2 | Gemini 2.0 Flash (Preview) | 61.8 |
| 3 | Gemini 1.5 Pro (002) | 48 |
| 4 | Gemini 1.5 Flash (002) | 45 |
| 5 | Gemma 3 27B (IT) | 40.8 |
| 6 | InternVL3-38B | 38.8 |
| 7 | InternVL2.5-78B | 38.6 |
| 8 | InternVL3-78B | 36.4 |
| 9 | GPT-4o | 36.2 |
| 10 | GPT-4o Mini | 35.6 |
| 11 | Gemma 3 12B (IT) | 33.8 |
| 12 | Qwen 2 VL 72B | 33.8 |
| 13 | Llama 4 Scout Base | 33.4 |
| 14 | Ovis2-8B | 32.4 |
| 15 | InternVL3-14B | 32 |
Interactive version: theaggregate.ai/benchmark?slug=mega-bench-task-scibench-with-solution-open-ended · How the rankings work · Data refreshed daily, snapshot 2026-07-22.