EuroEval Portuguese Common Sense Reasoning: leaderboard
Metric: Common Sense Reasoning Average Score (%). Source: euroeval.com. 395 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3 Pro (Preview) | 98.75 |
| 2 | Claude Sonnet 4.5 (Non-reasoning) | 98.65 |
| 3 | Claude Sonnet 4.5 (Thinking) | 98.45 |
| 4 | Gemma 4 31B (IT) | 97.87 |
| 5 | Claude Sonnet 4.6 | 96.68 |
| 6 | Gemini 2.5 Pro (Preview 03-25) | 96.07 |
| 7 | Qwen 3 235B A22B 2507 Instruct | 94.91 |
| 8 | GPT-4o (2024-05-13) | 94.53 |
| 9 | GPT-5 (Minimal) | 93.7 |
| 10 | Gemini 2.5 Pro | 93.37 |
| 11 | Mistral Small 3.1 | 93.34 |
| 12 | Grok 3 | 93.05 |
| 13 | Mistral Small 3 | 92.94 |
| 14 | Qwen 3 Next 80B A3B Instruct | 92.17 |
| 15 | Llama 3.1 405B Instruct FP8 | 92.02 |
Interactive version: theaggregate.ai/benchmark?slug=euroeval-portuguese-common-sense-reasoning · How It Works · Data refreshed daily, snapshot 2026-09-05.