Abliterated GLM 5.2 model, fine-tuned for security and red-team work

Abliterated GLM 5.2 is a fine-tuned language model with safety restrictions removed, designed for security research and red-team applications. It allows users to explore adversarial prompts, vulnerability analysis, and offensive security scenarios without standard content refusals. The model is accessible via a chat interface at abliteration.ai.