EuroEval Hungarian NLU - ScaLA HU — leaderboard
Metric: Linguistic acceptability Score (%). Source: euroeval.com. 203 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3 Pro (Preview) | 79.09 |
| 2 | Gemini 3 Flash (Preview) (Non-reasoning) | 74.15 |
| 3 | Gemini 2.5 Pro | 72.72 |
| 4 | GPT-4.1 | 70.5 |
| 5 | Gemini 3.1 Flash Lite (Preview) | 69.59 |
| 6 | GPT-5.4 Mini (Medium) | 68.79 |
| 7 | Claude Sonnet 4.5 (Thinking) | 68.18 |
| 8 | GPT-5.4 Mini (High) | 67.89 |
| 9 | Gemini 2.5 Flash (Thinking) | 67.25 |
| 10 | GPT-5.2 | 65.68 |
| 11 | GPT-5 | 64 |
| 12 | Claude Sonnet 4.6 | 63.91 |
| 13 | GPT-5 Mini | 60.51 |
| 14 | GPT-5.4 Nano (Medium) | 60.01 |
| 15 | Gemma 4 31B (IT) | 59.7 |
Interactive version: theaggregate.ai/benchmark?slug=euroeval-hungarian-nlu-scala-hu · How the rankings work · Data refreshed daily, snapshot 2026-07-22.