HELM Arabic - Mbzuai Human Translated Arabic Mmlu - World Religions: leaderboard
Metric: Exact Match (%). Source: crfm.stanford.edu. 41 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Sonnet 4.6 | 90.64 |
| 2 | Claude Opus 4.7 | 90.64 |
| 3 | GPT-5.4 (2026-03-05) | 89.47 |
| 4 | Gemma 4 31B (IT) | 87.72 |
| 5 | Gemini 2.5 Flash (Thinking) | 87.72 |
| 6 | Mistral Large 3 | 85.96 |
| 7 | GPT-4.1 (2025-04-14) | 84.21 |
| 8 | DeepSeek V3.1 | 84.21 |
| 9 | Qwen 3.5 397B A17B | 82.46 |
| 10 | Qwen 2.5 72B Instruct | 81.29 |
| 11 | Claude Haiku 4.5 (20251001) | 80.7 |
| 12 | AceGPT-v2-70B-Chat | 79.53 |
| 13 | Llama 3.3 70B Instruct | 78.95 |
| 14 | Mistral Large 2 (Nov) Instruct (2411) | 78.95 |
| 15 | Gemini 2.5 Flash Lite (Thinking) | 78.95 |
Interactive version: theaggregate.ai/benchmark?slug=helm-arabic-mbzuai-human-translated-arabic-mmlu-world-religions · How It Works · Data refreshed daily, snapshot 2026-09-19.