DarkKitsune/Qwen3.5-9B-Qworus-V2

VISIONPricing:Input $0.431 / Cached $0.0862 / Output $1.12Concurrent Unit Cost:1Model Size:9BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Aug 20, 2026Architecture:Transformer0.0K Featherless Exclusive Cold

DarkKitsune/Qwen3.5-9B-Qworus-V2 is a 9 billion parameter hybrid reasoning model, created by DarkKitsune through a 50/50 DARE-TIES merge of empero-ai/Qwen3.8-9B-Distill and ornith-ai/Ornith-1.5-9B. This model is specifically designed for complex tasks such as coding, tool use, planning, and general question answering. It offers specialized GGUF quantizations with an imatrix trained on a diverse mixture of code, business, math, and wikitext data, enhancing accuracy for longer contexts.

Loading preview...

Qwen3.5-9B-Qworus-V2: A Hybrid Reasoning Model

DarkKitsune/Qwen3.5-9B-Qworus-V2 is a 9 billion parameter language model developed by DarkKitsune. It is a 50/50 DARE-TIES merge of empero-ai/Qwen3.8-9B-Distill and ornith-ai/Ornith-1.5-9B, resulting in a robust hybrid architecture.

Key Capabilities

  • Advanced Reasoning: Engineered for complex reasoning tasks.
  • Coding & Tool Use: Excels in code generation, planning, and integrating with tools.
  • General Question Answering: Capable of providing comprehensive answers across various domains.
  • Optimized Quantizations: Provides GGUF quantizations (Q4_K_L, Q6_K_L) with an imatrix trained on a diverse dataset including code, business, math, and wikitext. These quantizations feature bumped-up token embed, output layers (Q8_0), and ssm_* tensors (BF16/Q8_0) for improved accuracy, especially with longer contexts.

Good For

  • Developers requiring a model for coding assistance and tool integration.
  • Applications needing strong planning and design capabilities.
  • Use cases demanding accurate reasoning and complex problem-solving.
  • Users seeking optimized local inference through specialized GGUF quantizations.