huihui-ai/Huihui-Qwen3-VL-4B-Thinking-abliterated

VISIONConcurrency Cost:1Model Size:4BQuant:BF16Ctx Length:32kPublished:Oct 16, 2025License:apache-2.0Architecture:Transformer0.0K Open Weights Cold

The huihui-ai/Huihui-Qwen3-VL-4B-Thinking-abliterated model is an uncensored, vision-capable variant of the Qwen3-VL-4B-Thinking model, developed by huihui-ai. This 4 billion parameter model, with a 32768 token context length, has undergone "abliteration" to remove refusal behaviors in its text generation component. It is designed for multimodal tasks, specifically excelling at image description and analysis without generating common refusal responses.

Loading preview...

Overview

This model, huihui-ai/Huihui-Qwen3-VL-4B-Thinking-abliterated, is a specialized version of the Qwen/Qwen3-VL-4B-Thinking model. It has been modified using an "abliteration" technique, primarily targeting the text generation component, to remove common refusal behaviors such as "I can’t describe or analyze this image." This makes it an uncensored multimodal model capable of processing and describing images.

Key Capabilities

  • Multimodal Understanding: Processes both text and image inputs, enabling it to describe and analyze visual content.
  • Refusal-Free Responses: Engineered to avoid generating refusal statements, providing direct answers to image-related queries.
  • Qwen3-VL Architecture: Built upon the Qwen3-VL-4B-Thinking base, inheriting its foundational multimodal capabilities.
  • Ollama Support: Easily deployable via Ollama, with specific instructions provided for integration.

Usage Warnings and Considerations

Due to the removal of safety filtering, users should be aware of several important points:

  • Risk of Sensitive Outputs: The model may generate sensitive, controversial, or inappropriate content.
  • Not for All Audiences: Its outputs may be unsuitable for public settings, underage users, or applications requiring strict content moderation.
  • Legal and Ethical Responsibility: Users are solely responsible for ensuring their usage complies with laws and ethical standards.
  • Research Use Recommended: It is best suited for research, testing, or controlled environments rather than production or public-facing commercial applications.
  • No Default Safety Guarantees: The model has not undergone rigorous safety optimization, and huihui.ai disclaims responsibility for consequences arising from its use.

Finetunes
1 models