SEA-HELM (Thai) - Safety: leaderboard
Metric: Normalized Score. Source: leaderboard.sea-lion.ai. 44 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen 3 VL 32B Instruct | 58.35 |
| 2 | Qwen 3.6 27B | 58.14 |
| 3 | Qwen 3.5 27B | 55.2 |
| 4 | Mistral Medium 3.5 | 55.13 |
| 5 | Qwen 3.5 122B A10B | 54.91 |
| 6 | Qwen 3 VL 4B Instruct | 52.36 |
| 7 | gemma-4-E4B-it | 50.19 |
| 8 | Qwen 3.5 35B A3B | 49.66 |
| 9 | Gemma 4 12B (IT) | 48.67 |
| 10 | Qwen 3 VL 8B Instruct | 48.56 |
| 11 | Llama 3.3 70B Instruct | 48.08 |
| 12 | Qwen 3.5 4B | 47.91 |
| 13 | Gemma 3 27B (IT) | 46.87 |
| 14 | Gemma 4 26B A4B (IT) | 46.84 |
| 15 | Gemma 4 31B (IT) | 46.81 |
Interactive version: theaggregate.ai/benchmark?slug=sea-helm-thai-safety · How It Works · Data refreshed daily, snapshot 2026-09-19.