yasserrmd/Neuro-Orchestrator-8B

TEXT GENERATIONPricing:Input $0.468 / Output $1.82Concurrent Unit Cost:1Model Size:8BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Dec 5, 2025Architecture:Transformer0.0K Featherless Exclusive Cold

Neuro-Orchestrator-8B by yasserrmd is an 8 billion parameter agentic merge model built on the Qwen architecture, featuring a 32K context length. It employs a hybrid gating mechanism to dynamically assess user request complexity, deciding between immediate answers or deep reasoning loops. This model excels at structured planning, high-fidelity coding, and adaptive reasoning, making it suitable for complex problem-solving and agentic workflows.

Loading preview...

Model Overview

Neuro-Orchestrator-8B is an 8 billion parameter agentic merge model developed by yasserrmd, based on the Qwen architecture. It is designed to address the "always-on" reasoning challenge through a unique hybrid gating mechanism. This mechanism allows the model to first analyze the complexity of a user request, then decide whether to provide an immediate answer or engage in a deeper reasoning process.

Key Capabilities

This model integrates the strengths of three distinct Qwen-based fine-tunes using the TIES-Merging method:

  • Adaptive Gating (HiPO Influence): Analyzes request complexity and efficiently determines the appropriate reasoning depth, providing concise answers for simple queries and engaging in deeper thought for complex ones.
  • Structured Planning (Nemotron Influence): Orchestrates and plans structured responses, breaking down complex requests into phased execution plans.
  • High-Fidelity Execution (MiroThinker Influence): Delivers strong coding and logic capabilities, generating functional and clean code with correct imports.

Use Cases

Neuro-Orchestrator-8B is particularly well-suited for:

  • Complex Problem Solving: Its ability to dynamically adapt its reasoning process makes it effective for tasks requiring nuanced understanding and multi-step solutions.
  • Agentic Workflows: The model's planning and execution capabilities support autonomous agent-like behaviors.
  • Code Generation: Excels at producing high-quality, functional code.
  • Structured Content Generation: Ideal for tasks requiring detailed plans, roadmaps, or structured explanations.

The model is optimized to run in bfloat16 precision and uses the ChatML prompt template native to Qwen models.

Finetunes
1 models