HELM Arabic Enterprise - Arabic Content Generation: leaderboard
The "Arabic Content Generation" subtask of HELM Arabic Enterprise. Stanford HELM's Arabic Enterprise project: business-facing Arabic tasks — content generation, summarisation, translation, dialect handling and customer-support style requests — scored by rubric and judge metrics rather than exact match.
Metric: Score. Source: crfm.stanford.edu. Saturation forecast: Around March 2027. 35 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemma 4 31B (IT) | 87.5 | #178 |
| 2 | GPT-5.4 (2026-03-05) | 86.41 | #58 |
| 3 | DeepSeek V4 Pro | 85.1 | #106 |
| 4 | GPT-5.4 Mini (2026-03-17) | 84.57 | #171 |
| 5 | Gemini 2.5 Flash (Non-reasoning) | 80.14 | #242 (Gemini 2.5 Flash) |
| 6 | Claude Opus 4.7 | 79.02 | #50 |
| 7 | Llama 4 Maverick Instruct | 78.34 | #460 |
| 8 | GPT-5.4 Nano (2026-03-17) | 77.4 | #266 |
| 9 | DeepSeek V3.1 | 77.36 | #268 |
| 10 | Gemini 2.5 Flash Lite (Non-reasoning) | 75.71 | #421 (Gemini 2.5 Flash Lite) |
| 11 | Claude Haiku 4.5 (20251001) | 75.68 | #258 |
| 12 | Qwen 3.5 9B | 73.76 | #384 |
| 13 | Llama 4 Scout Instruct | 72.64 | #542 |
| 14 | Claude Sonnet 4.6 | 72.37 | #92 |
| 15 | GPT-5.1 (2025-11-13) | 71.58 | #105 |
No result here: #4 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=helm-arabic-enterprise-arabic-content-generation · How It Works · Data refreshed daily, snapshot 2026-10-23.