wfen/Qwen3-VL-8B-Instruct-Heretic

VISIONConcurrent Unit Cost:1Model Size:8BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Nov 20, 2025License:apache-2.0Architecture:Transformer0.0K Open Weights Featherless Exclusive Cold

The wfen/Qwen3-VL-8B-Instruct-Heretic is an 8 billion parameter instruction-tuned multimodal large language model, based on the Qwen/Qwen3-VL-8B-Instruct architecture. This model is specifically fine-tuned for uncensored and 'heretic' responses, distinguishing it from its base model. It is designed for use cases requiring less restrictive content generation, leveraging its 32768 token context length for diverse applications.

Loading preview...

Model Overview

The wfen/Qwen3-VL-8B-Instruct-Heretic is an 8 billion parameter multimodal large language model, derived from the Qwen/Qwen3-VL-8B-Instruct base model. It has been instruction-tuned to provide responses that are less constrained by typical content filters, often referred to as 'uncensored' or 'heretic' in its design philosophy. This model supports a substantial context length of 32768 tokens, enabling it to process and generate longer, more complex interactions.

Key Characteristics

  • Base Model: Built upon the robust Qwen/Qwen3-VL-8B-Instruct architecture.
  • Parameter Count: Features 8 billion parameters, balancing performance with computational efficiency.
  • Context Length: Offers a 32768 token context window, suitable for extended dialogues and detailed inputs.
  • Fine-tuning Focus: Specifically fine-tuned on datasets like mlabonne/harmless_alpaca and mlabonne/harmful_behaviors to achieve its uncensored output style.
  • Multilingual Support: Supports both English and Chinese languages.

Use Cases

This model is particularly suited for applications where the generation of content without typical safety or ethical guardrails is desired. It can be used in research settings exploring model behavior with less restrictive outputs, or in creative applications where unfiltered text generation is a requirement. Developers should be aware of its 'uncensored' nature and use it responsibly, understanding the implications of its design.