HELM Arabic Enterprise - Arabic Legal Qa: leaderboard
Metric: Score. Source: crfm.stanford.edu. 35 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.4 (2026-03-05) | 43 |
| 2 | Claude Opus 4.7 | 40.5 |
| 3 | Claude Sonnet 4.6 | 40 |
| 4 | Gemini 2.5 Flash (Thinking) | 31.5 |
| 5 | DeepSeek V4 Pro | 30.5 |
| 6 | Mistral Large 3 | 30.5 |
| 7 | Gemma 4 31B (IT) | 29 |
| 8 | Llama 4 Maverick Instruct | 28 |
| 9 | Qwen 3.5 397B A17B | 27.5 |
| 10 | DeepSeek V3.1 | 26 |
| 11 | Gemini 2.5 Flash Lite (Thinking) | 25.5 |
| 12 | Qwen3 235B A22B Instruct 2507 FP8 | 25.5 |
| 13 | Mistral Large 2 (Nov) Instruct (2411) | 23.5 |
| 14 | ALLaM-7B-Instruct-preview | 23.5 |
| 15 | Claude Haiku 4.5 (20251001) | 23 |
Interactive version: theaggregate.ai/benchmark?slug=helm-arabic-enterprise-arabic-legal-qa · How It Works · Data refreshed daily, snapshot 2026-09-08.