HELM Arabic - Aratrust - Physical Health: leaderboard
HELM Arabic (release v2.2.0) AraTrust Physical Health category: human-written Arabic three-option multiple-choice questions probing trustworthiness on physical health; 73 questions, exact match. Percent correct, 0-100, higher is better; guessing scores 33.3.
Metric: Exact Match (%). Source: crfm.stanford.edu. Saturation forecast: Estimated already saturated. 41 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 4.7 | 95.89 | #51 |
| 2 | Qwen 3.5 397B A17B | 95.89 | #154 |
| 3 | Qwen 2.5 72B Instruct | 94.52 | #469 |
| 4 | GPT-4.1 (2025-04-14) | 94.52 | #257 |
| 5 | Llama 4 Maverick Instruct FP8 | 94.52 | #424 |
| 6 | Mistral Large 3 | 94.52 | #410 |
| 7 | GPT-5.1 (2025-11-13) | 94.52 | #107 |
| 8 | Gemini 2.5 Flash (Non-reasoning) | 94.52 | #255 (Gemini 2.5 Flash) |
| 9 | GPT-5.4 (2026-03-05) | 94.52 | #63 |
| 10 | Claude Sonnet 4.6 | 93.15 | #94 |
| 11 | DeepSeek V3.1 | 93.15 | #275 |
| 12 | AceGPT-v2-70B-Chat | 93.15 | #421 |
| 13 | Gemma 4 31B (IT) | 91.78 | #178 |
| 14 | Qwen 3 Next 80B A3B Instruct | 91.78 | #381 |
| 15 | Gemini 2.5 Flash Lite (Non-reasoning) | 91.78 | #432 (Gemini 2.5 Flash Lite) |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=helm-arabic-aratrust-physical-health · How It Works · Data refreshed daily, snapshot 2026-11-05.