Qwen3.8-27B-Abliterated-MTPLX-4bit

Model overview

What this model is for

Qwen3.8-27B-Abliterated-MTPLX-4bit is a 27-billion-parameter Hugging Face model intended for image-and-text input and image-text-to-text tasks. Its name explicitly identifies it as abliterated, and the directory record classifies it for vision-related use cases. This makes it potentially relevant for developers evaluating multimodal applications that combine visual content with text prompts, although the available source does not provide task examples, quality evaluations, implementation guidance, or inference benchmarks. Context-window size, tokens-per-second performance, and pricing are not disclosed. Licensing terms are also not disclosed in the supplied source, so review the Hugging Face repository and its files before using the model commercially or redistributing it. The available page extract primarily exposes the repository title and Hugging Face interface metadata; it does not provide additional technical specifications or a detailed model card. Consequently, compatibility with particular runtimes, quantization behavior, hardware requirements, and production safety characteristics cannot be confirmed here. The Hugging Face repository is the supplied source, but no publication or update date is available, so source freshness cannot be independently dated.

Classification

Categories, modalities and capabilities

Use-case categories

Input and output modalities

imagetext

Supported capabilities

abliterated

Continue exploring