HELM Arabic - Mbzuai Human Translated Arabic Mmlu - Econometrics: leaderboard
Metric: Exact Match (%). Source: crfm.stanford.edu. 41 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 4.7 | 80.7 |
| 2 | GPT-5.4 (2026-03-05) | 78.95 |
| 3 | Claude Sonnet 4.6 | 73.68 |
| 4 | Qwen 3.5 397B A17B | 72.81 |
| 5 | Gemma 4 31B (IT) | 67.54 |
| 6 | Claude Haiku 4.5 (20251001) | 64.91 |
| 7 | Llama 4 Maverick Instruct FP8 | 64.04 |
| 8 | DeepSeek V3.1 | 62.28 |
| 9 | Mistral Large 3 | 62.28 |
| 10 | Gemini 2.5 Flash (Thinking) | 62.28 |
| 11 | Qwen 3 Next 80B A3B Instruct | 61.4 |
| 12 | GPT-4.1 (2025-04-14) | 60.53 |
| 13 | Llama 4 Scout Instruct | 58.77 |
| 14 | GPT-4.1 Mini | 55.26 |
| 15 | Llama 3.3 70B Instruct | 53.51 |
Interactive version: theaggregate.ai/benchmark?slug=helm-arabic-mbzuai-human-translated-arabic-mmlu-econometrics · How It Works · Data refreshed daily, snapshot 2026-09-19.