Thai LLM MC - thaiexam_qa: leaderboard
The Thai LLM Leaderboard, run as an open results dataset by the Thai LLM community. Multiple-choice accuracy on ThaiExam, Thai national examination questions (SEACrowd harness).
Metric: Accuracy (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 70 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude 3.5 Sonnet (20240620) | 69.2 | #299 |
| 2 | GPT-4o (2024-05-13) | 63.89 | #325 |
| 3 | Qwen 2.5 32B Instruct | 61.42 | #498 |
| 4 | Qwen 2.5 72B Instruct | 60.53 | #454 |
| 5 | Llama 3.1 70B Cpt Sea Lionv3 Instruct | 60.53 | #548 |
| 6 | Llama 3.1 Nemotron 70B Instruct | 59.82 | #554 |
| 7 | Llama 3.3 70B Instruct | 59.29 | #534 |
| 8 | Llama 3.1 70B Instruct | 58.23 | #561 |
| 9 | Qwen 2 72B Instruct | 58.23 | #576 |
| 10 | Sailor2-20B-Chat | 58.05 | #590 |
| 11 | Qwen 2.5 14B Instruct | 57.35 | #647 |
| 12 | QwQ 32B-Preview | 57.35 | #651 |
| 13 | GPT-4o Mini (2024-07-18) | 54.51 | #558 |
| 14 | gemma2-9B-cpt-sea-lionv3-instruct | 54.51 | #705 |
| 15 | DeepSeek R1 Distill Qwen 32B | 54.34 | #655 |
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-mc-thaiexam-qa · How It Works · Data refreshed daily, snapshot 2026-10-20.