Babsie/Mistral-Nemo-Inst-2407-12B-Thinking-Uncensored-HERETIC-HI-Claude-Opus

Hugging Face
TEXT GENERATIONConcurrency Cost:1Model Size:12BQuant:FP8Ctx Length:32kPublished:May 21, 2026Architecture:Transformer0.0K Warm

Babsie/Mistral-Nemo-Inst-2407-12B-Thinking-Uncensored-HERETIC-HI-Claude-Opus is a 12 billion parameter Mistral Nemo Instruct model, fine-tuned by Babsie, featuring enhanced reasoning and thinking capabilities derived from Claude Opus 4.5. This model is specifically de-censored ('Heretic'd') to reduce refusal rates significantly, offering unfiltered and vivid prose for creative writing, roleplay, and complex problem-solving. It integrates a compact reasoning engine designed to improve output quality, detail, and complexity across a wide range of topics, including intense and visceral content.

Loading preview...

Model Overview

Babsie/Mistral-Nemo-Inst-2407-12B-Thinking-Uncensored-HERETIC-HI-Claude-Opus is a 12 billion parameter instruction-tuned model based on Mistral Nemo, developed by Babsie. It has been fine-tuned using an Unsloth process to integrate advanced reasoning and thinking capabilities, drawing inspiration from Claude Opus 4.5. A key differentiator is its 'Heretic' modification, which significantly reduces content refusal rates from 87/100 to 14/100, making it highly uncensored and capable of generating vivid, intense, and unfiltered content, including R-18 horror and explicit themes.

Key Capabilities

  • Enhanced Reasoning & Thinking: Incorporates a compact reasoning engine (300-600 tokens per reasoning block) that directly improves output detail, length, complexity, and overall quality without being long-winded.
  • Uncensored Output: Designed to provide unfiltered responses across various topics, including sensitive or explicit content, with a very low refusal rate.
  • Robust Performance: Reasoning activation is not affected by temperature settings (works from 0.1 to 2.5+), allowing for flexible generation styles.
  • Context Length: Supports a context window of 32768 tokens, with potential for up to 1 million tokens through its Claude Opus 4.5 integration.

Good For

  • Creative Writing & Roleplay: Excels at generating vivid, graphic, and intense narratives, particularly in genres like horror, with detailed descriptions and complex plots.
  • Unfiltered Content Generation: Ideal for applications requiring uncensored responses, humor, romance, or 'unfiltered truth' without built-in content restrictions.
  • Complex Problem Solving: The integrated thinking engine aids in generating more detailed and logically structured responses to intricate prompts.
  • Exploratory Generation: Users can experiment with a wide range of temperature settings to control output creativity without impacting the model's core reasoning function.

Finetunes
1 models