EuroEval Swedish NLU - SUC3: leaderboard
The "SUC3" subtask of EuroEval Swedish NLU. Swedish-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: Around January 2027. 408 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.4 Mini (2026-03-17) (High) | 83.44 | #169 (GPT-5.4 Mini (2026-03-17)) |
| 2 | Gemma 4 31B (IT) | 81.18 | #176 |
| 3 | Claude Sonnet 4.5 (Thinking) | 81 | #142 (Claude Sonnet 4.5) |
| 4 | GPT-5 | 80.47 | #98 |
| 5 | Gemma 4 26B A4B (IT) | 79.77 | #297 |
| 6 | GPT-5 (2025-08-07) (High) | 79.45 | |
| 7 | Qwen 3.5 35B A3B Base | 79.36 | #310 |
| 8 | Claude Sonnet 4.5 (Non-reasoning) | 78.37 | #142 (Claude Sonnet 4.5) |
| 9 | NVIDIA-Nemotron-3-Nano-30B-A3B-BF16 | 77.92 | #679 |
| 10 | Claude Sonnet 4.6 | 77.02 | #90 |
| 11 | Gemini 3 Pro (Preview) | 76.42 | #66 |
| 12 | Gemini 2.5 Flash (Thinking) | 76.3 | #245 (Gemini 2.5 Flash) |
| 13 | Llama 3.1 405B Instruct FP8 | 75.9 | #389 |
| 14 | GPT-5.2 | 75.83 | #110 |
| 15 | Grok 4.1 Fast (Reasoning) | 75.73 | #217 (Grok 4.1 Fast) |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=euroeval-swedish-nlu-suc3 · How It Works · Data refreshed daily, snapshot 2026-10-20.