Thai LLM - Roleplay: leaderboard
The Thai LLM Leaderboard, run as an open results dataset by the Thai LLM community. This board is the judge-rated 0-10 score for roleplay in Thai.
Metric: Rating (0-10). Source: huggingface.co. Saturation forecast: Estimated already saturated. 72 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 1.5 Pro (001) | 8 | #444 |
| 2 | GPT-4o (2024-05-13) | 7.9 | #325 |
| 3 | Gemini 1.5 Flash (001) | 7.85 | #577 |
| 4 | GPT-4o Mini (2024-07-18) | 7.8 | #558 |
| 5 | Llama 3.1 Nemotron 70B Instruct | 7.75 | #554 |
| 6 | QwQ 32B-Preview | 7.7 | #651 |
| 7 | gemma2-9B-cpt-sea-lionv3-instruct | 7.7 | #705 |
| 8 | Qwen 2.5 14B Instruct | 7.65 | #647 |
| 9 | DeepSeek R1 Distill Qwen 32B | 7.65 | #655 |
| 10 | Qwen 2.5 32B Instruct | 7.6 | #498 |
| 11 | Sailor2-20B-Chat | 7.6 | #590 |
| 12 | Qwen 2.5 72B Instruct | 7.55 | #454 |
| 13 | Gemma 2 9B (IT) | 7.5 | #788 |
| 14 | Qwen 2 72B Instruct | 7.5 | #576 |
| 15 | DeepSeek R1 Distill Llama 70B | 7.5 | #594 |
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-roleplay · How It Works · Data refreshed daily, snapshot 2026-10-20.