EVALITA - MAIA-GEN: leaderboard
The "MAIA-GEN" 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. 7 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemma 3 4B (IT) | 58.39 | #992 |
| 2 | Gemma 3 27B (IT) | 58.21 | #523 |
| 3 | Qwen 2.5 VL 32B Instruct | 55.94 | #453 |
| 4 | aya-vision-8B | 52.3 | #1107 |
| 5 | Qwen 2.5 VL 7B Instruct | 35.29 | #654 |
| 6 | Pixtral-12B | 32.42 | #808 |
| 7 | GLM-4.1V-9B (Thinking) | 9.86 | #472 (GLM-4.1V-9B) |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=evalita-maia-gen · How It Works · Data refreshed daily, snapshot 2026-10-20.