xtools-at/Qwen3.5-9B-Claude-Fable-5xMerge

VISIONPricing:Input $0.431 / Cached $0.0862 / Output $1.12Concurrent Unit Cost:1Model Size:9BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Oct 13, 2026License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Cold

xtools-at/Qwen3.5-9B-Claude-Fable-5xMerge is a 9 billion parameter experimental weighted merge of five Qwen3.5-9B fine-tunes, designed to combine agentic coding, reasoning distillation, and Claude Opus 4.6-style reasoning. This model integrates multi-turn software development, terminal workflows, structured problem-solving, and conversational generation. It aims to provide a compact model for complex tasks, leveraging a 32768 token context length for enhanced performance in coding, reasoning, and conversational AI applications.

Loading preview...

Qwen3.5-9B Claude Mythos/Fable/Opus 5xMerge Overview

This model is an experimental 9 billion parameter weighted merge of five distinct Qwen3.5-9B fine-tunes. It integrates various advanced capabilities by combining approaches from Claude Fable 5 agentic coding traces, Claude Mythos reasoning distillation, and Claude Opus 4.6-style reasoning fine-tuning. The merge aims to create a versatile checkpoint for complex AI tasks, building on the Qwen3.5-9B base architecture which supports a native context length of 262,144 tokens, though the merged model's effective context length is 32768 tokens.

Key Capabilities

  • Agentic Coding & Tool Use: Incorporates multi-turn software development, terminal workflows, and tool-use traces, including self-distilled coding behavior from Fable 5 agent trajectories.
  • Advanced Reasoning: Focuses on structured problem-solving across mathematics, science, and analytical tasks, with Claude-inspired thinking patterns and explicit intermediate reasoning.
  • Conversational Generation: Excels in general-purpose dialogue, instruction following, and extended text generation.
  • Long-Context Experimentation: Designed for tasks involving large documents or extended conversations, leveraging a lineage that includes long-context configurations.

Good For

  • Software Development: Ideal for code generation, debugging, code explanation, and multi-step implementation tasks.
  • Terminal & Agent Workflows: Suitable for structured tool calls, command-line task planning, and iterative task execution.
  • General Reasoning: Effective for analytical questions, mathematics, scientific reasoning, and structured problem-solving.
  • Conversational AI: Useful for dialogue, instruction following, and general-purpose text generation.
  • Experimental Applications: Provides a compact model for exploring combined reasoning-oriented generation and coding-agent capabilities.

Finetunes
1 models