EuroEval Portuguese NLU - ScaLA PT: leaderboard
Metric: Linguistic acceptability Score (%). Source: euroeval.com. 444 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Sonnet 4.6 | 76.33 |
| 2 | Gemini 3 Pro (Preview) | 74.07 |
| 3 | Gemini 3.1 Flash Lite (Preview) | 71.96 |
| 4 | GPT-5.2 | 71.49 |
| 5 | Gemma 4 31B (IT) | 71.04 |
| 6 | Gemma 4 26B A4B (IT) | 66.44 |
| 7 | GPT-5.4 Mini (Medium) | 65.51 |
| 8 | GPT-5.4 Mini (High) | 65.02 |
| 9 | GPT-5.4 Nano (High) | 64.73 |
| 10 | Grok 4.1 Fast (Reasoning) | 63.82 |
| 11 | GPT-5.4 Nano (Medium) | 61.35 |
| 12 | Llama 3.1 70B | 57.59 |
| 13 | Qwen 3.5 9B | 56.24 |
| 14 | gemma-3-27B-pt | 55.59 |
| 15 | Ministral 3 8B | 53.46 |
Interactive version: theaggregate.ai/benchmark?slug=euroeval-portuguese-nlu-scala-pt · How It Works · Data refreshed daily, snapshot 2026-09-05.