Open Portuguese LLM - HateBR Offensive — leaderboard
Metric: F1 (%). Source: huggingface.co. 42 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-4o (2024-08-06) | 93.2 |
| 2 | GPT-5 Mini (2025-08-07) | 93.06 |
| 3 | GPT-5 (2025-08-07) | 92.33 |
| 4 | DeepSeek V3.1 | 92.12 |
| 5 | GPT-5 Nano | 91.7 |
| 6 | Claude 3.7 Sonnet (20250219) | 91.25 |
| 7 | Gemini 1.5 Flash (001) | 90.92 |
| 8 | Llama 3.1 405B Instruct | 90.74 |
| 9 | DeepSeek V3 (0324) | 90.6 |
| 10 | Llama 4 Maverick Instruct | 90.48 |
| 11 | Gemini 1.5 Flash (002) | 90.46 |
| 12 | Llama 3.3 70B Instruct | 90.24 |
| 13 | Llama 3.2 90B Vision Instruct | 89.65 |
| 14 | Kimi K2 | 89.42 |
| 15 | Gemini 2.5 Flash | 89.33 |
Interactive version: theaggregate.ai/benchmark?slug=open-portuguese-llm-hatebr-offensive · How the rankings work · Data refreshed daily, snapshot 2026-07-22.