HELM Arabic - Mbzuai Human Translated Arabic Mmlu - Sociology: leaderboard
Metric: Exact Match (%). Source: crfm.stanford.edu. 41 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 4.7 | 92.04 |
| 2 | Mistral Large 3 | 89.05 |
| 3 | GPT-5.4 (2026-03-05) | 89.05 |
| 4 | Gemini 2.5 Flash (Thinking) | 88.56 |
| 5 | Qwen 3.5 397B A17B | 88.06 |
| 6 | Claude Sonnet 4.6 | 87.56 |
| 7 | Claude Haiku 4.5 (20251001) | 87.06 |
| 8 | GPT-4.1 (2025-04-14) | 85.57 |
| 9 | Gemma 4 31B (IT) | 84.58 |
| 10 | Qwen 2.5 72B Instruct | 84.08 |
| 11 | Llama 4 Maverick Instruct FP8 | 84.08 |
| 12 | DeepSeek V3.1 | 83.58 |
| 13 | Gemini 2.5 Flash Lite (Thinking) | 83.08 |
| 14 | Llama 3.3 70B Instruct | 81.09 |
| 15 | Llama 4 Scout Instruct | 80.1 |
Interactive version: theaggregate.ai/benchmark?slug=helm-arabic-mbzuai-human-translated-arabic-mmlu-sociology · How It Works · Data refreshed daily, snapshot 2026-09-19.