HELM Arabic - Arabic Mmlu - Computer Science (Middle School): leaderboard
HELM Arabic (release v2.2.0) ArabicMMLU Computer Science (Middle School): native Arabic multiple-choice questions from ArabicMMLU, which draws on school, university and professional exams across Arabic-speaking countries; 27 questions, exact match. Percent correct, 0-100, higher is better.
Metric: Exact Match (%). Source: crfm.stanford.edu. Saturation forecast: Estimated already saturated. 41 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemma 4 31B (IT) | 100 | #178 |
| 2 | Llama 4 Scout Instruct | 100 | #547 |
| 3 | GPT-4.1 (2025-04-14) | 100 | #257 |
| 4 | Mistral Large 2 (Nov) Instruct (2411) | 100 | #480 |
| 5 | GPT-5.4 Mini (2026-03-17) | 100 | #180 |
| 6 | Llama 3.3 70B Instruct | 96.3 | #541 |
| 7 | GPT-4.1 Mini | 96.3 | #363 |
| 8 | Qwen 3.5 397B A17B | 96.3 | #154 |
| 9 | Mistral Large 3 | 96.3 | #410 |
| 10 | Gemini 2.5 Flash (Non-reasoning) | 96.3 | #255 (Gemini 2.5 Flash) |
| 11 | Qwen 2.5 72B Instruct | 92.59 | #469 |
| 12 | Claude Opus 4.7 | 92.59 | #51 |
| 13 | Claude Haiku 4.5 (20251001) | 92.59 | #264 |
| 14 | Qwen 3 Next 80B A3B Instruct | 92.59 | #381 |
| 15 | DeepSeek V3.1 | 92.59 | #275 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=helm-arabic-arabic-mmlu-computer-science-middle-school · How It Works · Data refreshed daily, snapshot 2026-11-05.