Open Italian LLM - MMLU-Pro (IT): leaderboard
Italian MMLU-Pro from the Open Italian LLM Leaderboard: a harder multiple-choice knowledge-and-reasoning benchmark translated to Italian, reporting accuracy.
Metric: Accuracy (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 29 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Llama 3 8B Ita | 30.7 | #926 |
| 2 | LLaMAntino-3-ANITA-8B-Inst-DPO-ITA | 30.11 | #1009 |
| 3 | Llama 3 8B Instruct | 28.43 | #1129 |
| 4 | Llama 3 8B 4bit UltraChat Ita | 26.01 | #1280 |
| 5 | Mistral-7B-v0.1 | 25.04 | #1326 |
| 6 | Mistral-7B-v0.3 | 24.9 | #1355 |
| 7 | Volare | 23.61 | #1213 |
| 8 | Qwen 2 1.5B Instruct | 20.93 | #1426 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=open-italian-llm-mmlu-pro-it · How It Works · Data refreshed daily, snapshot 2026-10-20.