bbsai/Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NM-DAU

Hugging Face
VISIONPricing:Input $1.6 / Cached $0.15 / Output $12Concurrent Unit Cost:2Model Size:27BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Sep 19, 2026License:apache-2.0Architecture:Transformer0.0K Open Weights Featherless Exclusive Warm

The bbsai/Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NM-DAU model, developed by DavidAU, is a 27 billion parameter Qwen 3.8-based language model with a 32768 token context length. It is fine-tuned for enhanced reasoning, reduced thinking token usage, and uncensored output, achieving an ARC-C score of 0.735, significantly higher than the base Qwen 3.8-27B-Instruct. This model excels in complex analytical tasks and creative generation, maintaining high performance even at 4-bit quantization.

Loading preview...

Model Overview

The bbsai/Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NM-DAU is a 27 billion parameter model built upon the Qwen 3.8 architecture, developed by DavidAU. This particular release, designated as "Release #1," is part of a series of highly optimized Qwen variants. It features a substantial context length of 32768 tokens, making it suitable for processing extensive inputs.

Key Capabilities & Differentiators

  • Enhanced Reasoning: Achieves an ARC-C score of 0.735, which is 141 points higher than the base Qwen 3.8 27B-Instruct model, indicating superior reasoning capabilities.
  • Efficient Thinking: Incorporates a "thinking token" reduction mechanism, using 1/2 to 1/10 the tokens of standard Qwen models while maintaining detail and quality.
  • Uncensored Output: Undergoes a "Heretic'ing" process to remove safety alignments, resulting in uncensored and unfiltered responses, balanced with a low KLD (Kullback-Leibler Divergence) for minimal performance impact.
  • Quantization Performance: Demonstrates robust performance at 4-bit quantization, retaining approximately 99% of 8-bit performance.
  • Creative & Analytical Strength: Optimized for both detailed analytical tasks and creative content generation, as evidenced by example generations.

Ideal Use Cases

  • Complex Problem Solving: Suited for applications requiring advanced reasoning and analytical depth.
  • Creative Writing & Roleplay: Its uncensored nature and creative enhancements make it strong for generating diverse and unrestricted narratives.
  • Research & Development: Useful for exploring model behavior without inherent safety biases, particularly in research settings where unfiltered output is desired.

Finetunes
1 models