HELM Arabic Enterprise - Arabic Finance Calculation: leaderboard
The "Arabic Finance Calculation" 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: Accuracy. Source: crfm.stanford.edu. Saturation forecast: Around April 2027. 35 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 4.7 | 75.34 | #50 |
| 2 | Claude Sonnet 4.6 | 64.84 | #92 |
| 3 | DeepSeek V4 Pro | 63.47 | #106 |
| 4 | GPT-5.4 (2026-03-05) | 62.56 | #58 |
| 5 | Llama 4 Maverick Instruct | 60.27 | #460 |
| 6 | Gemma 4 31B (IT) | 59.36 | #178 |
| 7 | DeepSeek V3.1 | 57.99 | #268 |
| 8 | GPT-5.1 (2025-11-13) | 57.08 | #105 |
| 9 | Qwen 3.5 397B A17B | 55.25 | #147 |
| 10 | Gemini 2.5 Flash (Non-reasoning) | 54.79 | #242 (Gemini 2.5 Flash) |
| 11 | Mistral Large 3 | 53.88 | #404 |
| 12 | GPT-5.4 Mini (2026-03-17) | 53.42 | #171 |
| 13 | Claude Haiku 4.5 (20251001) | 52.51 | #258 |
| 14 | Gemini 2.5 Flash Lite (Non-reasoning) | 50.68 | #421 (Gemini 2.5 Flash Lite) |
| 15 | Qwen 3.5 9B | 49.77 | #384 |
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-finance-calculation · How It Works · Data refreshed daily, snapshot 2026-10-23.