HELM ThaiExam - ThaiExam: leaderboard
HELM ThaiExam aggregate score across Thai standardized-exam subjects.
Metric: EM. Source: crfm.stanford.edu. Saturation forecast: Estimated already saturated. 42 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude 3.5 Sonnet (20240620) | 75.1 | #299 |
| 2 | Claude 3 Opus (20240229) | 68.57 | #412 |
| 3 | Llama 3.1 405B Instruct | 68.23 | #457 |
| 4 | Gemini 1.5 Pro (001) | 64.7 | #444 |
| 5 | Qwen 2 72B Instruct | 63.78 | #576 |
| 6 | GPT-4o (2024-05-13) | 63.26 | #325 |
| 7 | Llama 3.1 70B Instruct | 62.33 | #561 |
| 8 | Gemini 1.5 Flash (001) | 59.8 | #577 |
| 9 | Claude 3 Sonnet (20240229) | 57.93 | #677 |
| 10 | GPT-4 Turbo | 57.48 | #404 |
| 11 | Llama 3 70B Instruct | 56.83 | #634 |
| 12 | GPT-4o Mini (2024-07-18) | 55.74 | #558 |
| 13 | Claude 3 Haiku (20240307) | 53.95 | #771 |
| 14 | Qwen 1.5 72B Chat | 53.75 | #776 |
| 15 | SeaLLM-7B-v2.5 | 47.84 | #1071 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=helm-thaiexam-thaiexam · How It Works · Data refreshed daily, snapshot 2026-10-20.