DarkArtsForge/Morax-24B-v2

Morax 24B v2: Uncensored Creative Narrative Generation

Morax 24B v2, developed by DarkArtsForge, is a 24 billion parameter language model specifically engineered for creative, intelligent, and uncensored content generation, including graphic and erotic narratives. This model is a merge of Darkhn/Magistral-2509-24B-Text-Only, Naphula/BeaverAI_Fallen-Mistral-Small-3.1-24B-v1e_textonly, and TheDrummer/Precog-24B-v1, utilizing the della merge method.

Key Capabilities & Characteristics

• Uncensored Output: Designed with normalize: false to ensure no refusals or need for ablation/jailbreaks, even for sensitive or explicit prompts.
• Enhanced Performance: Achieved a significantly higher score of 10351 on Q0 Bench compared to v1's 7656, indicating improved overall capability.
• "Evil-Tuned" for Specific Content: Optimized for generating "deranged and depraved" narratives and roleplay, including violent and graphic erotic content.
• Dynamic Thinking Mechanism: The model can optionally engage in an internal "thinking" process, which can be controlled by users with <think></nothink> tags.
• Robust Context Handling: Demonstrates no early terminations with thinking mode, even on prompts that typically trigger EOS in other models like Naphula-Archives/T13-24B-GGUF.

Ideal Use Cases

• Creative Writing & Roleplay: Particularly suited for generating explicit, dark, or unrestricted fictional content.
• Uncensored Content Generation: For applications requiring a model that will not refuse prompts based on content filters.
• Experimental Narrative Development: Users exploring the boundaries of AI-generated storytelling without typical safety constraints.