DavidAU/Qwen3.6-27B-Fable-Fusion-711-Uncensored-Heretic-NM-DAU-MTP

Hugging Face
VISIONConcurrent Unit Cost:2Model Size:27BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Jul 28, 2026License:apache-2.0Architecture:Transformer0.1K Open Weights Featherless Exclusive Warm

DavidAU/Qwen3.6-27B-Fable-Fusion-711-Uncensored-Heretic-NM-DAU-MTP is a 27 billion parameter Qwen3.6-based multi-stage fine-tuned and merged model, developed by a collaboration including DavidAU and Nightmedia. It is notable for being the first model of its size to exceed 700 ARC-C in both 8-bit and 4-bit quantization, outperforming the base Qwen3.6-27B and Qwen3.6-35B-A3B across most benchmarks. This model is designed to enhance general intelligence, problem-solving, and instruction following, making it suitable for a wide range of general-purpose applications on consumer hardware.

Loading preview...

Overview

DavidAU/Qwen3.6-27B-Fable-Fusion-711 is a 27 billion parameter model built on the Qwen3.6 architecture, resulting from a multi-stage fine-tuning and merging process by DavidAU, Nightmedia, and other collaborators. It is specifically optimized for consumer hardware and is the first model of its size to achieve an ARC-C score exceeding 700 in both 8-bit and 4-bit quantizations, surpassing the performance of the base Qwen3.6-27B and Qwen3.6-35B-A3B models across 6 out of 7 benchmarks.

Key Capabilities

  • Enhanced Intelligence & Problem Solving: Significantly improves general model intelligence, thinking, and reasoning abilities.
  • Superior Instruction Following: Designed for better adherence to complex instructions.
  • Vision Activated: Supports multimodal inputs, requiring a separate mmproj file for image processing.
  • Uncensored: A de-censored version, offering broader content generation capabilities.
  • Optimized Quantization: Utilizes NEO IMATRIX GGUF quants, improving accuracy by 2-4% and enhancing long context performance, with MTP (multi-token prediction) GGUFs available for increased speed.
  • Extended Context: Natively supports a 256k context length, extensible up to 1,010,000 tokens using YaRN scaling.

Good For

  • General-purpose AI applications: Excels in a broad range of tasks due to improved intelligence and instruction following.
  • Consumer Hardware Deployment: Optimized for efficient operation on typical user systems.
  • Applications requiring high accuracy and reasoning: Particularly in scenarios where precise problem-solving is critical.
  • Multimodal tasks: With its activated vision capabilities, it can process image inputs.
  • Developers seeking uncensored model behavior: Offers flexibility in content generation.

Finetunes
1 models