MEGA-Bench Task - Code Execution: leaderboard
Metric: Task Score (%). Source: huggingface.co. 44 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.5 Pro (Preview 03-25) | 93.8 |
| 2 | GPT-4o | 75 |
| 3 | InternVL3-78B | 50 |
| 4 | GPT-4o Mini | 50 |
| 5 | Claude 3.5 Sonnet (20240620) | 50 |
| 6 | Gemma 3 27B (IT) | 43.8 |
| 7 | Gemini 1.5 Pro (002) | 43.8 |
| 8 | Gemini 1.5 Flash (002) | 43.8 |
| 9 | InternVL2.5-78B | 43.8 |
| 10 | Claude 3.5 Sonnet (20241022) | 37.5 |
| 11 | Gemma 3 12B (IT) | 25 |
| 12 | InternVL2-8B | 25 |
| 13 | Qwen 2 VL 72B | 25 |
| 14 | InternVL3-38B | 25 |
| 15 | Llama 4 Scout Base | 25 |
Interactive version: theaggregate.ai/benchmark?slug=mega-bench-task-code-execution · How It Works · Data refreshed daily, snapshot 2026-09-05.