HELM Arabic - Mbzuai Human Translated Arabic Mmlu - Security Studies: leaderboard
Metric: Exact Match (%). Source: crfm.stanford.edu. 41 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Sonnet 4.6 | 80.82 |
| 2 | GPT-5.4 (2026-03-05) | 80.82 |
| 3 | Gemma 4 31B (IT) | 80.41 |
| 4 | Claude Opus 4.7 | 80 |
| 5 | Qwen 3.5 397B A17B | 78.37 |
| 6 | Llama 4 Maverick Instruct FP8 | 76.73 |
| 7 | Mistral Large 3 | 75.92 |
| 8 | Mistral Large 2 (Nov) Instruct (2411) | 75.92 |
| 9 | Claude Haiku 4.5 (20251001) | 75.51 |
| 10 | Gemini 2.5 Flash (Thinking) | 75.51 |
| 11 | DeepSeek V3.1 | 75.1 |
| 12 | Llama 3.3 70B Instruct | 73.88 |
| 13 | GPT-4.1 (2025-04-14) | 73.88 |
| 14 | Llama 4 Scout Instruct | 73.47 |
| 15 | Gemini 2.5 Flash Lite (Thinking) | 73.47 |
Interactive version: theaggregate.ai/benchmark?slug=helm-arabic-mbzuai-human-translated-arabic-mmlu-security-studies · How It Works · Data refreshed daily, snapshot 2026-09-19.