SeaEval - Fundamental NLP Tasks - C3 (Zero-Shot): leaderboard
Metric: Accuracy (%). Source: huggingface.co. 25 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Llama 3.1 70B Cpt Sea Lionv3 Instruct | 96.41 |
| 2 | Qwen 2 72B Instruct | 96.11 |
| 3 | Llama 3.1 70B Instruct | 96.04 |
| 4 | Qwen 2.5 32B Instruct | 96.04 |
| 5 | Qwen 2.5 72B Instruct | 95.96 |
| 6 | Llama 3 70B Instruct | 95.21 |
| 7 | Qwen 2.5 14B Instruct | 95.03 |
| 8 | gemma2-9B-cpt-sea-lionv3-instruct | 92.56 |
| 9 | Qwen 2 7B Instruct | 92.45 |
| 10 | Gemma 2 9B (IT) | 92.22 |
| 11 | SeaLLMs-v3-7B-Chat | 91.44 |
| 12 | Qwen 2.5 7B Instruct | 91.21 |
| 13 | Sailor2-8B-Chat | 89.6 |
| 14 | Llama 3.1 8B Instruct | 88.15 |
| 15 | Llama 3 8B Cpt Sea Lionv2.1 Instruct | 86.76 |
Interactive version: theaggregate.ai/benchmark?slug=seaeval-fundamental-nlp-tasks-c3-zero-shot · How It Works · Data refreshed daily, snapshot 2026-09-05.