A fine-tuned variant of Meta's Llama 3.1 8B with safety filters removed, this model responds to prompts that standard models typically decline. It operates with a large 128K context window and handles text-in, text-out tasks. The trade-off is clear: without built-in guardrails, outputs require careful handling and responsible deployment by the user.