HELM Arabic - Arabic Mmlu - Computer Science (Primary School): leaderboard
Metric: Exact Match (%). Source: crfm.stanford.edu. 41 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 4.7 | 90 |
| 2 | GPT-4.1 (2025-04-14) | 86.84 |
| 3 | Llama 4 Maverick Instruct FP8 | 85.79 |
| 4 | Llama 4 Scout Instruct | 85.26 |
| 5 | Gemma 4 31B (IT) | 85.26 |
| 6 | Gemini 2.5 Flash (Thinking) | 85.26 |
| 7 | Claude Sonnet 4.6 | 84.74 |
| 8 | GPT-5.4 (2026-03-05) | 84.74 |
| 9 | Gemini 2.5 Flash Lite (Thinking) | 83.16 |
| 10 | GPT-4.1 Mini | 82.11 |
| 11 | Qwen 3 Next 80B A3B Instruct | 82.11 |
| 12 | Qwen 3.5 397B A17B | 81.05 |
| 13 | Mistral Large 3 | 80.53 |
| 14 | Claude Haiku 4.5 (20251001) | 79.47 |
| 15 | GPT-4.1 Nano | 77.89 |
Interactive version: theaggregate.ai/benchmark?slug=helm-arabic-arabic-mmlu-computer-science-primary-school · How It Works · Data refreshed daily, snapshot 2026-09-19.