cyberneurova/cyberneurova-Qwen3.8-27B

Hugging Face
VISIONPricing:Input $1.6 / Cached $0.15 / Output $12Concurrent Unit Cost:2Model Size:27BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Aug 15, 2026License:apache-2.0Architecture:Transformer0.0K Open Weights Featherless Exclusive Warm

The cyberneurova-Qwen3.8-27B model, developed by CyberNeurova, is a 27 billion parameter, multimodal, uncensored, and neutral assistant built upon Qwen/Qwen3.8-27B. It features a 262K context window, multilingual support, and tool-calling capabilities. This model is designed to provide direct answers without unsolicited disclaimers and excels in reasoning and executable coding tasks, achieving 92% pass@1 on an executable coding benchmark.

Loading preview...

Overview

cyberneurova-Qwen3.8-27B is a 27 billion parameter, multimodal language model developed by CyberNeurova, based on the Qwen/Qwen3.8-27B architecture. It is designed to be an uncensored and neutral assistant, providing direct answers and fully honoring system prompts without merging in default behaviors. The model supports a substantial 262K context window, is multilingual, and includes tool-calling functionality.

Key Capabilities

  • Uncensored and Steerable: Provides direct responses and allows complete control over tone and policy via system prompts, without built-in refusals or unsolicited disclaimers.
  • Multimodal (Vision): Capable of processing images, allowing users to include image URLs or base64 encoded images in prompts for visual understanding tasks.
  • Strong Reasoning and Coding Performance: Demonstrates high capability in reasoning tasks and achieves a 92% pass@1 on an executable coding benchmark, indicating proficiency in generating functional code.
  • Flexible Deployment: Can be served in BF16 and runs out-of-the-box with vLLM, transformers, or any OpenAI-compatible server, with GGUF builds available for local inference via llama.cpp/Ollama.

Good For

  • Applications requiring a highly steerable and neutral assistant that adheres strictly to provided system prompts.
  • Tasks involving complex reasoning and problem-solving, including executable code generation.
  • Multimodal applications that require image understanding alongside text processing.
  • Developers seeking a model with preserved capabilities for general knowledge and direct responses, suitable for a wide range of conversational and analytical use cases.