EuroEval Italian NLU: leaderboard
Italian-language EuroEval natural-language-understanding suite, separating NLU task performance from the broader all-task EuroEval aggregate.
Metric: NLU Average Score (%). Source: euroeval.com. Saturation forecast: Around May 2027. 435 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Pro (Preview) | 75.73 | #66 |
| 2 | Gemma 4 31B (IT) | 75.1 | #176 |
| 3 | Claude Sonnet 4.6 | 73.35 | #90 |
| 4 | Gemini 3.1 Flash Lite (Preview) | 73.27 | #194 |
| 5 | Llama 3.1 70B | 71.9 | #592 |
| 6 | GPT-5.4 Mini (2026-03-17) (High) | 71.16 | #169 (GPT-5.4 Mini (2026-03-17)) |
| 7 | Gemini 3 Flash (Preview) (Thinking) | 71.1 | #82 (Gemini 3 Flash (Preview)) |
| 8 | Gemma 4 26B A4B (IT) | 70.68 | #297 |
| 9 | Qwen 3.5 35B A3B Base | 70.26 | #310 |
| 10 | Gemini 3 Flash (Preview) (Non-reasoning) | 70.08 | #82 (Gemini 3 Flash (Preview)) |
| 11 | Qwen 3 235B A22B 2507 FP8 (Thinking) | 67.91 | #246 (Qwen 3 235B A22B 2507 FP8) |
| 12 | Grok 4.1 Fast (Reasoning) | 67.47 | #217 (Grok 4.1 Fast) |
| 13 | NVIDIA-Nemotron-3-Nano-30B-A3B-BF16 | 67.36 | #679 |
| 14 | GLM-4.7 Flash | 66.72 | #509 |
| 15 | GPT-5.2 | 66.5 | #110 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=euroeval-italian-nlu · How It Works · Data refreshed daily, snapshot 2026-10-20.