nightmedia/Qwen3.5-9B-Holodeck-Tess-Bradbury-Orwell

VISIONConcurrent Unit Cost:1Model Size:9BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Jul 20, 2026License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Cold

The nightmedia/Qwen3.5-9B-Holodeck-Tess-Bradbury-Orwell model is a 9 billion parameter language model created by nightmedia, formed by merging several Qwen3.5-9B variants and migtissera/Tess-4-9B. This merged model demonstrates improved performance across various reasoning and common sense benchmarks compared to its baseline components. It is designed for general language understanding and generation tasks, offering a balanced performance profile for diverse applications.

Loading preview...

Model Overview

nightmedia/Qwen3.5-9B-Holodeck-Tess-Bradbury-Orwell is a 9 billion parameter language model developed by nightmedia through a sophisticated merging process. It combines the strengths of migtissera/Tess-4-9B, Qwen3.5-9B-TNG-PKD-Qwopus-Coder-Fable-Polaris, and Qwen3.5-9B-TNG-PKD-Qwopus-Fable-Polaris-Bradbury-Orwell-Polaris using the nuslerp merge method.

Key Characteristics

  • Merged Architecture: Built upon the Qwen3.5-9B family, integrating multiple specialized models to enhance overall capabilities.
  • Improved Benchmarks: Demonstrates enhanced performance across a suite of benchmarks including ARC, BoolQ, HSwag, OBQA, PIQA, and WinoGrande, showing improvements over the Qwen3.5-9B-Instruct baseline.
  • Quantization Performance: Provides detailed perplexity and tokens/sec metrics across various quantization levels (e.g., mxfp8, q8-hi, q6-hi, q4-hi), indicating efficiency and performance trade-offs for deployment.

When to Use This Model

  • General-Purpose Applications: Suitable for a wide range of natural language processing tasks where a balanced performance across reasoning and common sense is beneficial.
  • Resource-Aware Deployment: The provided quantization metrics allow developers to select optimal configurations for specific hardware constraints, balancing performance and memory usage.
  • Exploration of Merged Models: Ideal for users interested in leveraging the combined strengths of multiple base models for potentially more robust and versatile language generation.