EuroEval Serbian NLU - UNER SR: leaderboard
The "UNER 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: Micro F1 excluding MISC (%). Source: euroeval.com. Saturation forecast: Estimated already saturated. 205 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.4 Mini (2026-03-17) (High) | 88.77 | #169 (GPT-5.4 Mini (2026-03-17)) |
| 2 | Claude Sonnet 4.5 (Thinking) | 87.6 | #142 (Claude Sonnet 4.5) |
| 3 | GPT-5 | 87.13 | #98 |
| 4 | Qwen 3.5 35B A3B Base | 85.81 | #310 |
| 5 | Claude Sonnet 4.5 (Non-reasoning) | 85.69 | #142 (Claude Sonnet 4.5) |
| 6 | Qwen 3 4B 2507 (Thinking) | 85.21 | #538 (Qwen 3 4B 2507) |
| 7 | Gemini 3 Flash (Preview) (Thinking) | 84.81 | #82 (Gemini 3 Flash (Preview)) |
| 8 | Gemini 2.5 Flash (Thinking) | 84.58 | #245 (Gemini 2.5 Flash) |
| 9 | Gemma 4 31B (IT) | 84.1 | #176 |
| 10 | Gemini 3 Flash (Preview) (Non-reasoning) | 83.15 | #82 (Gemini 3 Flash (Preview)) |
| 11 | Llama 3.3 70B Instruct | 82.88 | #534 |
| 12 | Gemma 4 26B A4B (IT) | 81.73 | #297 |
| 13 | GPT-5 Nano | 81.22 | #425 |
| 14 | Mistral Small 3.1 | 81.05 | #610 |
| 15 | Claude Sonnet 4.6 | 80.97 | #90 |
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-uner-sr · How It Works · Data refreshed daily, snapshot 2026-10-20.