HELM Arabic - Mbzuai Human Translated Arabic Mmlu - International Law: leaderboard
Metric: Exact Match (%). Source: crfm.stanford.edu. 41 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 4.7 | 92.56 |
| 2 | GPT-5.4 (2026-03-05) | 90.91 |
| 3 | GPT-4.1 (2025-04-14) | 89.26 |
| 4 | Gemini 2.5 Flash (Thinking) | 89.26 |
| 5 | Claude Sonnet 4.6 | 88.43 |
| 6 | Llama 4 Maverick Instruct FP8 | 87.6 |
| 7 | Qwen 3.5 397B A17B | 87.6 |
| 8 | Gemma 4 31B (IT) | 85.95 |
| 9 | Mistral Large 3 | 85.12 |
| 10 | Llama 3.3 70B Instruct | 83.47 |
| 11 | Qwen 3 Next 80B A3B Instruct | 83.47 |
| 12 | DeepSeek V3.1 | 83.47 |
| 13 | Claude Haiku 4.5 (20251001) | 82.64 |
| 14 | AceGPT-v2-70B-Chat | 81.82 |
| 15 | Qwen 2.5 72B Instruct | 80.17 |
Interactive version: theaggregate.ai/benchmark?slug=helm-arabic-mbzuai-human-translated-arabic-mmlu-international-law · How It Works · Data refreshed daily, snapshot 2026-09-19.