EuroEval Lithuanian NLU: leaderboard
Lithuanian-language EuroEval natural-language-understanding suite, separating NLU task performance from the broader all-task EuroEval aggregate.
Metric: NLU Average Score (%). Source: euroeval.com. Saturation forecast: Estimated already saturated. 235 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Pro (Preview) | 59.01 | #66 |
| 2 | Qwen 3 235B A22B 2507 Instruct | 58.89 | #300 |
| 3 | Qwen 3 Next 80B A3B Instruct | 56.63 | #368 |
| 4 | gemma-3-27B-pt | 56.32 | #502 |
| 5 | Claude Sonnet 4.6 | 56.1 | #90 |
| 6 | Magistral Small 1.2 | 55.98 | #570 |
| 7 | Llama 3.1 70B | 55.55 | #592 |
| 8 | Gemini 3 Flash (Preview) (Non-reasoning) | 55.45 | #82 (Gemini 3 Flash (Preview)) |
| 9 | Qwen 3 235B A22B 2507 FP8 (Thinking) | 55.17 | #246 (Qwen 3 235B A22B 2507 FP8) |
| 10 | Qwen 3 14B | 55.05 | #537 |
| 11 | Gemini 3.1 Flash Lite (Preview) | 54.99 | #194 |
| 12 | Mistral Small 3.2 | 54.87 | #605 |
| 13 | Gemini 3 Flash (Preview) (Thinking) | 54.85 | #82 (Gemini 3 Flash (Preview)) |
| 14 | GPT-5 | 54.84 | #98 |
| 15 | Gemini 2.5 Pro | 54.82 | #152 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=euroeval-lithuanian-nlu · How It Works · Data refreshed daily, snapshot 2026-10-20.