Greek MMLU: leaderboard
Greek MMLU — Greek-language multitask knowledge exam adapted from MMLU; accuracy (%).
Metric: Accuracy (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 84 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Flash (Preview) | 93.16 | #82 |
| 2 | GPT-5.2 | 87.75 | #110 |
| 3 | GPT-4o (2024-11-20) | 86.81 | #381 |
| 4 | Qwen 2.5 72B Instruct | 79.7 | #454 |
| 5 | Llama 3.3 70B Instruct | 79.65 | #534 |
| 6 | Gemma 3 27B (IT) | 79.41 | #523 |
| 7 | Qwen 2.5 72B | 79.2 | #499 |
| 8 | Llama 3.1 70B Instruct | 79.13 | #561 |
| 9 | gemma-3-27B-pt | 78.88 | #502 |
| 10 | Qwen 3 30B A3B 2507 Instruct | 78.39 | #488 |
| 11 | Llama 3.1 70B | 75.71 | #592 |
| 12 | Gemma 3 12B (IT) | 75.31 | #664 |
| 13 | gemma-3-12B-pt | 74.99 | #686 |
| 14 | Qwen 2.5 32B Instruct | 73.22 | #498 |
| 15 | Qwen 2.5 32B | 73.14 | #546 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=greek-mmlu · How It Works · Data refreshed daily, snapshot 2026-10-20.