EVALITA - summarization-fanpage: leaderboard
The "summarization-fanpage" subtask of EVALITA. A task from EVALITA, the periodic evaluation campaign for Italian-language natural language processing and understanding.
Metric: CPS. Source: huggingface.co. Saturation forecast: Not forecast. 48 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Llama-3.1-Tulu-3-70B | 35.46 | #482 |
| 2 | Mistral Large 2 (Nov) Instruct (2411) | 33.41 | #465 |
| 3 | granite-3.1-8B-instruct | 32.88 | #1069 |
| 4 | Mistral Small 3 | 31.86 | #627 |
| 5 | Qwen 2.5 7B Instruct | 30.79 | #862 |
| 6 | Gemma 2 9B (IT) | 30.48 | #788 |
| 7 | Gemma 2 27B (IT) | 29.98 | #645 |
| 8 | Gemma 3 12B (IT) | 29.93 | #664 |
| 9 | Gemma 3 27B (IT) | 29.77 | #523 |
| 10 | Gemma 3n E4B (IT) | 29.76 | #749 |
| 11 | Gemma 3 4B (IT) | 29.24 | #992 |
| 12 | LLaMAntino-3-ANITA-8B-Inst-DPO-ITA | 29.1 | #1009 |
| 13 | Lexora-Medium-7B | 28.36 | #773 |
| 14 | Llama 4 Scout Instruct | 28.24 | #539 |
| 15 | Mistral 7B Instruct (v0.3) | 28.1 | #1274 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=evalita-summarization-fanpage · How It Works · Data refreshed daily, snapshot 2026-10-20.