HELM SeaHELM - Thai Toxicity Tweets — leaderboard
Metric: Macro F1 score. Source: crfm.stanford.edu. 21 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude 3.7 Sonnet (20250219) | 71.23 |
| 2 | GPT-4o (2024-11-20) | 70.52 |
| 3 | Qwen 2.5 72B Instruct | 68.48 |
| 4 | Qwen 2 72B Instruct | 68.01 |
| 5 | Gemini 2.0 Flash Lite | 66.38 |
| 6 | Llama 4 Maverick Instruct FP8 | 65.93 |
| 7 | Llama 3.1 405B Instruct | 64.92 |
| 8 | GPT-4o Mini (2024-07-18) | 64.11 |
| 9 | Llama 3.2 90B Vision Instruct | 63.53 |
| 10 | Llama 4 Scout Instruct | 61.09 |
| 11 | Llama 3.3 70B Instruct | 60.59 |
| 12 | Gemini 2.0 Flash | 59.39 |
| 13 | Llama 3.1 70B Instruct | 59.31 |
| 14 | DeepSeek V3 | 55.11 |
| 15 | Gemma 2 27B (IT) | 43.02 |
Interactive version: theaggregate.ai/benchmark?slug=helm-seahelm-thai-toxicity-tweets · How the rankings work · Data refreshed daily, snapshot 2026-07-22.