Open Arabic LLM - Arabic MMLU HT Machine Learning Leaderboard — The Aggregate

Open Arabic LLM - Arabic MMLU HT Machine Learning — leaderboard

Metric: Accuracy (%). Source: huggingface.co. 163 models tracked.

Top models

#	Model	Score
1	calme-2.1-qwen2.5-72B	60.71
2	Rombos-LLM-V2.5-Qwen-72B	60.71
3	Ultiima-72B	60.71
4	calme-2.2-qwen2.5-72B	58.93
5	Qwen 2.5 72B Instruct	56.25
6	Qwentile2.5-32B-Instruct	56.25
7	Qwen 2.5 72B	53.57
8	Qwen 2 72B	53.57
9	lambda-qwen2.5-32B-dpo-test	53.57
10	Awqward2.5-32B-Instruct	52.68
11	Llama 3.3 70B Instruct	50.89
12	Qwen2.5-32B-Instruct-CFT	50.89
13	Qwen 2.5 32B Instruct	50
14	Qwen 2.5 32B	49.11
15	Rombos-LLM-V2.6-Qwen-14B	49.11

Interactive version: theaggregate.ai/benchmark?slug=open-arabic-llm-arabic-mmlu-ht-machine-learning · How the rankings work · Data refreshed daily, snapshot 2026-07-22.