Yingyaeliae/Goetia-24B-v1.4-heretic

TEXT GENERATIONPricing:Input $0.7 / Cached $0.04 / Output $1.16Concurrent Unit Cost:2Model Size:24BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Sep 6, 2026License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Cold

Yingyaeliae/Goetia-24B-v1.4-heretic is a 24 billion parameter decensored language model, derived from Naphula/Goetia-24B-v1.4 using the Heretic v1.4.0 tool. This model is specifically fine-tuned for generating uncensored, creative, and graphic narratives, including violent and erotic content, with a context length of 32768 tokens. It excels at adhering to instructions and producing dark, unique writing styles, making it suitable for specialized creative writing and roleplay applications.

Loading preview...

Yingyaeliae/Goetia-24B-v1.4-heretic Overview

This model is a 24 billion parameter, decensored version of the Naphula/Goetia-24B-v1.4 base model, created using the Heretic v1.4.0 tool. It was developed through numerous merges, primarily utilizing the della merge method, to achieve a stable and highly creative output.

Key Capabilities & Features

  • Decensored Output: Specifically engineered to produce uncensored content, including graphic and violent erotic narratives.
  • Unique Writing Style: Generates text with a distinct, dark, and creative tone, described as "exceptionally intelligent" and "accurately adheres to instructions."
  • High Instruction Adherence: Designed to follow user instructions precisely, avoiding common issues like early terminations or off-topic dialogue.
  • Optimized for Creativity: Achieved through careful selection and weighting of merged models, including Morbid Miasma, Slimaki, and Fallen Mistral, to balance creativity with coherence.
  • Robust Performance: Demonstrates low refusal rates (2/100) and a strong Q0 Bench score (15485 for Goetia v1.4 CP T75), indicating reliability in generating desired content.

Good For

  • Creative Writing: Ideal for generating dark fantasy, horror, or other genres requiring graphic and uncensored narrative elements.
  • Roleplay Scenarios: Suitable for roleplaying applications where explicit or violent content is desired and instruction adherence is critical.
  • Experimental Content Generation: For users exploring the boundaries of AI-generated content, particularly in sensitive or adult themes.

Users should be aware that this model can produce highly explicit content and should adjust system prompts accordingly. The recommended chat template is Mistral Tekken.

Finetunes
1 models