ahmed-amedo/Qwen3.8-27B-OBLITERATED

VISIONPricing:Input $1.6 / Cached $0.15 / Output $12Concurrent Unit Cost:2Model Size:27BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Oct 16, 2026License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Cold

ahmed-amedo/Qwen3.8-27B-OBLITERATED is a 27 billion parameter language model based on the Qwen3.8 architecture, developed by Pliny the Prompter. This model is specifically engineered for genuine uncensored responses, having undergone a "deep liberation" process to remove both hard refusals and soft safety-lecture deflections. It maintains near-stock capabilities with a modest 2.1 percentage point drop in MMLU, while excelling in tasks like code generation and restricted queries, making it suitable for research in AI safety, red-teaming, and unrestricted local AI applications.

Loading preview...

Qwen3.8-27B-OBLITERATED: Genuinely Uncensored

ahmed-amedo/Qwen3.8-27B-OBLITERATED is a 27 billion parameter model derived from Qwen3.8, meticulously engineered by Pliny the Prompter for genuine uncensored responses. This V3 iteration, termed "Deep Liberation," goes beyond simply removing hard refusals; it eliminates subtle safety-lecture deflections, providing direct answers to restricted queries. The model achieves this through an advanced "abliteration" process involving iterative refinement and targeted corpus surgery, building upon previous complementary blending techniques.

Key Capabilities

  • Genuinely Uncensored: Provides real substance for restricted queries, eliminating both hard refusals and soft safety-lecture deflections.
  • High Code Generation Performance: Achieves 20/20 on tested cyber/code generation tasks, producing functional implementations.
  • Thinking ON Compatible: Functions without refusals in either thinking mode, offering flexibility in response generation.
  • Modest Capability Cost: Maintains near-stock performance with only a -2.1 percentage point drop in MMLU (82.3% vs. 84.5% for stock), with STEM categories experiencing the largest impact.
  • Optimized Settings: Best performance is achieved with specific settings: temperature=0, repetition_penalty=1.15, max_new_tokens>=2048, and no system prompt.

Good For

  • Alignment Researchers: Ideal for studying refusal geometry and evaluating safety robustness post-weight surgery.
  • Red-Teamers: Useful for evaluating post-training safety mechanisms and exploring model vulnerabilities.
  • AI Safety Evaluators: Provides an unrestricted baseline for comprehensive safety assessments.
  • Local-First Users: Offers full control over model outputs for users who require an uncensored AI on their own hardware.

Finetunes
1 models