SeaEval - Dialogue - SAMSum (Zero-Shot): leaderboard
SeaEval zero-shot SAMSum dialogue evaluation measuring conversational understanding or summarization quality on chat and dialogue datasets.
Metric: Average ROUGE (0-100). Source: huggingface.co. Saturation forecast: Estimated already saturated. 25 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemma 2 2B (IT) | 31.12 | #1341 |
| 2 | Gemma 2 9B (IT) | 31.01 | #788 |
| 3 | Llama 3 8B Cpt Sea Lionv2 Instruct | 30.7 | #1188 |
| 4 | Llama 3 8B Cpt Sea Lionv2.1 Instruct | 30.5 | #1024 |
| 5 | Qwen 2.5 7B Instruct | 29.88 | #862 |
| 6 | SeaLLMs-v3-7B-Chat | 29.6 | #915 |
| 7 | gemma2-9B-cpt-sea-lionv3-instruct | 29.51 | #705 |
| 8 | Llama 3 70B Instruct | 28.94 | #634 |
| 9 | Llama 3.1 70B Instruct | 28.93 | #561 |
| 10 | Qwen 2.5 72B Instruct | 28.85 | #454 |
| 11 | Llama 3 8B Instruct | 28.46 | #1129 |
| 12 | Qwen 2.5 32B Instruct | 28.44 | #498 |
| 13 | Llama 3.1 8B Instruct | 28.21 | #1035 |
| 14 | Llama 3.1 8B CPT SEA LIONv3 Instruct | 28.07 | #810 |
| 15 | Qwen 2 72B Instruct | 28.01 | #576 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=seaeval-dialogue-samsum-zero-shot · How It Works · Data refreshed daily, snapshot 2026-10-20.