HELM Arabic - Arabic Mmlu - Accounting (University): leaderboard
Metric: Exact Match (%). Source: crfm.stanford.edu. 41 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 4.7 | 89.19 |
| 2 | Claude Sonnet 4.6 | 87.84 |
| 3 | Gemma 4 31B (IT) | 87.84 |
| 4 | GPT-5.4 (2026-03-05) | 87.84 |
| 5 | Qwen 3.5 397B A17B | 86.49 |
| 6 | Claude Haiku 4.5 (20251001) | 85.14 |
| 7 | Gemini 2.5 Flash (Thinking) | 81.08 |
| 8 | Gemini 2.5 Flash Lite (Thinking) | 81.08 |
| 9 | GPT-4.1 (2025-04-14) | 79.73 |
| 10 | Mistral Large 3 | 79.73 |
| 11 | Llama 4 Maverick Instruct FP8 | 75.68 |
| 12 | DeepSeek V3.1 | 75.68 |
| 13 | Llama 4 Scout Instruct | 74.32 |
| 14 | Qwen 3.5 9B | 74.32 |
| 15 | GPT-4.1 Mini | 72.97 |
Interactive version: theaggregate.ai/benchmark?slug=helm-arabic-arabic-mmlu-accounting-university · How It Works · Data refreshed daily, snapshot 2026-09-19.