EuroEval Finnish NLU - ScaLA FI: leaderboard
The "ScaLA FI" subtask of EuroEval Finnish NLU. Finnish-language EuroEval natural-language-understanding suite, separating NLU task performance from the broader all-task EuroEval aggregate.
Metric: Matthews correlation coefficient (%). Source: euroeval.com. Saturation forecast: Around January 2028. 372 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Pro (Preview) | 62.11 | #66 |
| 2 | Gemini 3.1 Flash Lite (Preview) | 59.08 | #194 |
| 3 | Gemma 4 31B (IT) | 58.36 | #176 |
| 4 | Gemini 3 Flash (Preview) (Non-reasoning) | 54.95 | #82 (Gemini 3 Flash (Preview)) |
| 5 | Gemini 2.5 Pro | 53.35 | #152 |
| 6 | GPT-5.4 Nano (2026-03-17) (High) | 51.4 | #266 (GPT-5.4 Nano (2026-03-17)) |
| 7 | Gemini 3 Flash (Preview) (Thinking) | 51.06 | #82 (Gemini 3 Flash (Preview)) |
| 8 | GPT-5 (2025-08-07) (High) | 48.12 | |
| 9 | Gemini 2.5 Flash (Preview 04-17) | 47.25 | #237 |
| 10 | O3 (2025-04-16) | 46.79 | #121 |
| 11 | GPT-5 | 46.59 | #98 |
| 12 | Qwen 3 235B A22B 2507 Instruct | 45.86 | #300 |
| 13 | Gemma 4 26B A4B (IT) | 45.41 | #297 |
| 14 | GPT-4.1 Mini | 45.18 | #355 |
| 15 | Gemini 2.5 Flash (Thinking) | 44.49 | #245 (Gemini 2.5 Flash) |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=euroeval-finnish-nlu-scala-fi · How It Works · Data refreshed daily, snapshot 2026-10-20.