Greek MMLU - Humanities — leaderboard

Metric: Accuracy (%). Source: huggingface.co. 62 models tracked.

Top models

#ModelScore
1Gemini 3 Flash (Preview)92.88
2GPT-4o (2024-11-20)88.68
3GPT-5.288.27
4Llama 3.1 70B Instruct82.34
5Llama 3.3 70B Instruct82.2
6Gemma 3 27B (IT)79.87
7gemma-3-27B-pt79.83
8Qwen 2.5 72B Instruct79.14
9Llama 3.1 70B79.05
10Qwen 2.5 72B78.78
11EuroLLM-22B-Instruct-251275.31
12Gemma 3 12B (IT)75.13
13Qwen 3 30B A3B 2507 Instruct74.81
14gemma-3-12B-pt74.44
15Mistral-Small-24B-Base-250174.07

Interactive version: aibenchmarks.dev/benchmark?slug=greek-mmlu-humanities · How the rankings work · Data refreshed daily, snapshot 2026-07-08.