Thai LLM NLG - iapp_squad_seacrowd_qa (ROUGEL): leaderboard
The Thai LLM Leaderboard, run as an open results dataset by the Thai LLM community. Generation quality on iapp-SQuAD Thai extractive question answering, scored by ROUGE-L (SEACrowd harness).
Metric: ROUGEL. Source: huggingface.co. Saturation forecast: Estimated already saturated. 68 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Llama 3 8B Cpt Sea Lionv2.1 Instruct | 74.2 | #1024 |
| 2 | Llama 3.1 70B Cpt Sea Lionv3 Instruct | 71.45 | #548 |
| 3 | Llama 3.3 70B Instruct | 71.07 | #534 |
| 4 | Llama 3.1 8B Instruct | 68.57 | #1035 |
| 5 | Llama 3.1 70B Instruct | 64.63 | #561 |
| 6 | Llama 3.1 8B CPT SEA LIONv3 Instruct | 62.8 | #810 |
| 7 | Llama 3.1 Nemotron 70B Instruct | 61.11 | #554 |
| 8 | Llama 3.2 3B Instruct | 59.89 | #1352 |
| 9 | llama-3-typhoon-v1.5-8B-instruct | 55.99 | #1198 |
| 10 | Llama 3 8B Cpt Sea Lionv2 Instruct | 53.61 | #1188 |
| 11 | SeaLLMs-v3-7B-Chat | 48.12 | #915 |
| 12 | Llama 3.2 1B Instruct | 45.77 | #1539 |
| 13 | gemma2-9B-cpt-sea-lionv3-instruct | 38.74 | #705 |
| 14 | GPT-4o Mini (2024-07-18) | 38.17 | #558 |
| 15 | Qwen 2 7B Instruct | 37.25 | #991 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=thai-llm-nlg-iapp-squad-seacrowd-qa-rougel · How It Works · Data refreshed daily, snapshot 2026-10-20.