HELM Arabic - Mbzuai Human Translated Arabic Mmlu - High School Government And Politics: leaderboard
Metric: Exact Match (%). Source: crfm.stanford.edu. 41 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Sonnet 4.6 | 96.89 |
| 2 | Claude Opus 4.7 | 96.37 |
| 3 | Gemini 2.5 Flash (Thinking) | 96.37 |
| 4 | Qwen 3.5 397B A17B | 95.34 |
| 5 | GPT-5.4 (2026-03-05) | 95.34 |
| 6 | Claude Haiku 4.5 (20251001) | 92.75 |
| 7 | GPT-4.1 (2025-04-14) | 91.19 |
| 8 | AceGPT-v2-70B-Chat | 90.67 |
| 9 | Gemma 4 31B (IT) | 90.16 |
| 10 | Mistral Large 3 | 90.16 |
| 11 | DeepSeek V3.1 | 89.12 |
| 12 | Qwen 2.5 72B Instruct | 88.6 |
| 13 | Llama 4 Maverick Instruct FP8 | 87.56 |
| 14 | Llama 3.3 70B Instruct | 86.53 |
| 15 | Qwen 3 Next 80B A3B Instruct | 86.53 |
Interactive version: theaggregate.ai/benchmark?slug=helm-arabic-mbzuai-human-translated-arabic-mmlu-high-school-government-and-politics · How It Works · Data refreshed daily, snapshot 2026-09-19.