DarkArtsForge/Magistaroth-24B-v1.1

Hugging Face
TEXT GENERATIONConcurrency Cost:2Model Size:24BQuant:FP8Ctx Length:32kPublished:Mar 1, 2026License:apache-2.0Architecture:Transformer0.0K Open Weights Warm

Magistaroth-24B-v1.1 by DarkArtsForge is a 24 billion parameter language model with a 32768 token context length, specifically designed for highly creative, detailed, and uncensored narrative generation and roleplay. It utilizes a custom 'pdq' merge method to enhance novelty and explore unique vector regions, resulting in a model optimized for imaginative content. While it may rarely exhibit minor artifacts like missing spaces or misspellings due to its aggressive merge technique, its primary use case is generating unrestricted and vivid textual content.

Loading preview...

Magistaroth 24B v1.1 Overview

Magistaroth 24B v1.1 is a 24 billion parameter language model developed by DarkArtsForge, distinguished by its focus on generating highly creative, detailed, and uncensored narratives and roleplay content. It features a substantial 32768 token context length, allowing for extensive and coherent story development.

Key Capabilities

  • Enhanced Creativity: Utilizes a unique pdq merge method that re-scans donor models to explore 'dark matter' regions of vectors, synergistically augmenting the base model with novel and unique outputs.
  • Uncensored Content Generation: Specifically configured to be fully uncensored, allowing for the creation of narratives and roleplay scenarios that may include violent and graphic erotic content, as noted by the developer.
  • Detailed Narrative & Roleplay: Optimized for producing rich, imaginative, and in-depth textual content, making it suitable for complex storytelling and interactive roleplaying.
  • High Performance: Achieves a Q0 bench score of 12648, indicating strong performance in its specialized domain.

Good for

  • Creative Writing: Ideal for authors and writers seeking a model that can generate highly imaginative and detailed stories without content restrictions.
  • Roleplaying Scenarios: Excellent for users who require a model capable of engaging in complex, uncensored, and vivid roleplay interactions.
  • Exploring Novel Text Generation: Suitable for researchers and developers interested in pushing the boundaries of transformer creativity through unique merging techniques.

Important Considerations

  • Content Warning: Users should be aware that this model can produce violent and graphic erotic content. Adjusting the system prompt accordingly is recommended.
  • Potential Artifacts: Due to the aggressive pdq merge method, there is a small risk of rare artifacts such as missing spaces, misspelled, or repeated words. These are generally minor and can be edited out.
  • Recommended Settings: For optimal performance, the developer recommends using temp 1.0, topnsigma 1.25, and the Mistral Tekken chat template.

Finetunes
2 models