ZeroXClem/Qwen3.5-9B-Fable-5-Quad-Stock

VISIONConcurrency Cost:1Model Size:9BQuant:FP8Ctx Length:32kTool Calling:SupportedPublished:Jul 3, 2026License:apache-2.0Architecture:Transformer0.0K Open Weights Cold

ZeroXClem/Qwen3.5-9B-Fable-5-Quad-Stock is a 9.41 billion parameter multimodal language model based on the Qwen3.5-9B architecture. It is a four-way fusion of popular Claude Fable 5 distills, created using MergeKit's Model Stock method. This model excels in Fable-flavored chat, reasoning, code generation, UI design, and instruction following, while also supporting image input through its retained vision tower.

Loading preview...

Overview

ZeroXClem/Qwen3.5-9B-Fable-5-Quad-Stock is a 9.41 billion parameter model built on the Qwen3.5-9B base. It's a unique four-way fusion of the most-downloaded Claude Fable 5 distills, merged using MergeKit's Model Stock method. This approach geometrically recenters four specialized Fable-flavored models—focused on reasoning, coding, UI design, and instruction-following—around the Qwen3.5-9B base.

Key Capabilities

  • Multimodal Input: Fully supports image input, as the complete Qwen3.5 vision tower was retained and merged. This allows for image description, analysis, and reasoning.
  • Fable-flavored Responses: Delivers coherent, structured responses with a distinct "storyteller's voice" derived from its Claude Fable 5 ancestry.
  • Code Generation & UI Design: Incorporates specialized training for code generation, review, and frontend prototyping, including component-structured UI thinking.
  • Instruction Following: Excels at instruction following and generating structured outputs due to its SFT-Glint ancestry.
  • Performance: Benchmarks show it maintains the base model's reasoning capabilities, with ARC-Challenge scores comparable to the Qwen3.5-9B base.

Good For

  • General Chat & Reasoning: Ideal for applications requiring nuanced, structured conversational AI.
  • Software Development: Suitable for code generation, review, and UI design tasks.
  • Multimodal Applications: Excellent for scenarios requiring image understanding alongside text-based interactions.
  • Structured Output Generation: Effective for tasks needing precise instruction following and formatted responses.

Finetunes
9 models