EuroEval English NLU - SST-5: leaderboard
Metric: Sentiment classification Score (%). Source: euroeval.com. 335 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Sonnet 4.6 | 71.56 |
| 2 | GPT-4.1 | 71 |
| 3 | GPT-5 | 70.36 |
| 4 | Llama 3.1 405B Instruct FP8 | 70.29 |
| 5 | Claude Sonnet 4.5 (Non-reasoning) | 70.16 |
| 6 | QwQ-32B | 70.14 |
| 7 | Qwen 3 Next 80B A3B Instruct | 69.82 |
| 8 | gemma-3-27B-pt | 69.57 |
| 9 | GPT-5.4 Mini (Medium) | 69.49 |
| 10 | Yi 1.5 34B | 69.38 |
| 11 | GLM-4.1V-9B (Thinking) | 69.38 |
| 12 | Qwen 3 235B A22B 2507 Instruct | 69.36 |
| 13 | Gemma 3 27B (IT) | 69.29 |
| 14 | Gemma 2 9B (IT) | 69.27 |
| 15 | Gemini 3.1 Flash Lite (Preview) | 69.2 |
Interactive version: theaggregate.ai/benchmark?slug=euroeval-english-nlu-sst-5 · How It Works · Data refreshed daily, snapshot 2026-09-05.