Classifies English text into hate speech, offensive, or neither. Fine-tuned on 110K+ examples from Davidson, ImplicitHate, HateXplain, and HateDay 2025.