SAHM - Open-Ended QA Average: leaderboard
Metric: Mean judge score (0-10) over the three open-ended tasks (event-cause QA, Islamic standards QA, fatwa QA) of SAHM, an Arabic financial and Sharia-compliance benchmark, each answer graded against the gold answer by a Gemini 2.5 Flash judge with an additive rubric, greedy decoding, mean of three runs; higher is better. Source: arxiv.org. Saturation forecast: Around January 2027. 20 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 4.5 | 8.84 |
| 2 | GPT-5 | 8.84 |
| 3 | Claude Sonnet 4.5 | 8.4 |
| 4 | Claude Haiku 4.5 | 7.55 |
| 5 | GPT-4o | 7.16 |
| 6 | Gemma 3 27B (IT) | 6.69 |
| 7 | Qwen 2.5 72B Instruct | 6.37 |
| 8 | GPT-4o Mini | 6.29 |
| 9 | Fanar-1-9B-Instruct | 5.67 |
| 10 | Qwen 2.5 14B Instruct | 5.48 |
| 11 | ALLaM-7B-Instruct-preview | 5.35 |
| 12 | Llama 3.1 70B Instruct | 5.04 |
| 13 | Gemma 2 9B (IT) | 4.93 |
| 14 | Gemma 3 4B (IT) | 4.26 |
| 15 | Qwen 2.5 7B Instruct | 4.06 |
Interactive version: theaggregate.ai/benchmark?slug=sahm-open-ended-qa-average · How It Works · Data refreshed daily, snapshot 2026-10-07.