Naphula/Goetia-31B-v1
Naphula/Goetia-31B-v1 is a 31 billion parameter Gemma4-based language model merged using the della method, featuring a 32768 token context length. This model is designed to be fully uncensored, producing narratives and roleplay content without refusals, and is optimized for creative and graphic content generation. It excels in scenarios requiring unrestricted text output and is recommended for use with the Gemma 4 chat template.
Loading preview...
Goetia 31B v1 Overview
Naphula/Goetia-31B-v1 is a 31 billion parameter language model built on the Gemma4 architecture, utilizing a 32768 token context length. It was created using the della merge method, combining multiple Gemma4-31B base models to achieve its unique characteristics. The model is specifically engineered to be fully uncensored, meaning it does not exhibit refusals during basic Q0 benchmark tests and is capable of generating a wide range of content without built-in restrictions.
Key Capabilities
- Uncensored Content Generation: Designed to produce narratives and roleplay content, including potentially violent and graphic erotic material, without inherent censorship or refusal mechanisms.
- Merge Architecture: Leverages the
dellamerge method, integrating contributions from various Gemma4-31B models like ApocalypseParty, AuriAetherwiing, ConicCat, and DavidAU, among others. - High Context Length: Supports a substantial context window of 32768 tokens, allowing for more extensive and coherent interactions.
Good For
- Creative Writing & Roleplay: Ideal for applications requiring unrestricted and imaginative text generation, particularly for narratives that may include sensitive or graphic themes.
- Experimental AI Development: Suitable for developers and researchers exploring the boundaries of language model capabilities without content filters.
Users should be aware of the model's uncensored nature and adjust system prompts accordingly, with the Gemma 4 template recommended for optimal performance.