EuroEval Serbian NLU - MultiWikiQA SR: leaderboard
The "MultiWikiQA SR" subtask of EuroEval Serbian NLU. Serbian-language EuroEval natural-language-understanding suite, separating NLU task performance from the broader all-task EuroEval aggregate.
Metric: Character F1 (%). Source: euroeval.com. Saturation forecast: Around January 2027. 205 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Ministral-3-14B-Reasoning-2512 | 68.07 | #540 |
| 2 | Mistral Small 3.2 | 67.16 | #605 |
| 3 | gemma-3-27B-pt | 66.09 | #502 |
| 4 | Llama 3.1 70B | 65.97 | #592 |
| 5 | GPT-5 Mini (2025-08-07) | 65.59 | #172 |
| 6 | Qwen 3 14B (Non-reasoning) | 65.41 | #537 (Qwen 3 14B) |
| 7 | Qwen3.5-4B-Base | 65.38 | #798 |
| 8 | GPT-5.2 | 65 | #110 |
| 9 | Gemini 3 Pro (Preview) | 64.13 | #66 |
| 10 | SOLAR-10.7B-v1.0 | 64.1 | #1253 |
| 11 | Grok 4.1 Fast (Reasoning) | 63.97 | #217 (Grok 4.1 Fast) |
| 12 | Qwen3.5-2B-Base | 63.45 | #1309 |
| 13 | Claude Sonnet 4.5 (Non-reasoning) | 63.43 | #142 (Claude Sonnet 4.5) |
| 14 | Qwen 3.5 35B A3B Base | 63.36 | #310 |
| 15 | Gemini 2.5 Flash (Thinking) | 63.24 | #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-serbian-nlu-multiwikiqa-sr · How It Works · Data refreshed daily, snapshot 2026-10-20.