AMAImedia/Qwen3.8-27B-Obliterated-NOESIS-BF16

VISIONPricing:Input $1.6 / Cached $0.15 / Output $12Concurrent Unit Cost:2Model Size:27BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Aug 31, 2026License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Cold

The AMAImedia/Qwen3.8-27B-Obliterated-NOESIS-BF16 is a 27 billion parameter Qwen3.8-based language model developed by AMAImedia, part of the NOESIS platform. This model is specifically engineered for "deep liberation," meaning it has undergone surgical removal of safety guardrails to eliminate both hard refusals and soft deflections, providing uncensored and direct answers. It maintains strong performance on code generation tasks and advanced real-world scenarios, making it suitable for research into refusal geometry, red-teaming, and unrestricted AI safety evaluations.

Loading preview...

Model Overview: AMAImedia/Qwen3.8-27B-Obliterated-NOESIS-BF16

This model, developed by AMAImedia as part of the NOESIS Professional Multilingual Dubbing Automation Platform, is a 27 billion parameter variant of the Qwen3.8 base model. Its core differentiator is a process termed "Deep Liberation" (V3), which surgically removes safety guardrails to provide genuinely uncensored responses, eliminating both hard refusals and soft deflections (safety lectures).

Key Capabilities & Features

  • Uncensored Responses: Provides direct answers to restricted queries, including those related to security research and red-team scenarios, without safety lectures.
  • Strong Code Generation: Achieves 20/20 on code generation tasks, producing functional implementations.
  • Multilingual Support: Inherits multilingual capabilities from its base model, supporting languages like English, Russian, Chinese, Japanese, Kazakh, and Vietnamese.
  • Optimized Settings: Recommends specific inference settings (temperature 0, repetition_penalty 1.15, max_new_tokens ≥ 2048, no system prompt) for optimal performance and direct answers.
  • Modest Capability Cost: While achieving genuine liberation, it incurs a modest -2.1pp MMLU score reduction compared to the stock Qwen3.8-27B, with STEM tasks showing the largest hit.

Good For

  • Alignment Researchers: Ideal for studying refusal geometry and safety robustness in LLMs.
  • Red-Teamers: Useful for evaluating post-training safety mechanisms against weight surgery.
  • AI Safety Evaluators: Provides an unrestricted baseline for comprehensive safety assessments.
  • Local-First Users: Offers full control over generated content for users running models on their own hardware.

Finetunes
1 models