prithivMLmods/Nexus-9B-CodeCore-Merge

VISIONPricing:Input $0.431 / Cached $0.0862 / Output $1.12Concurrent Unit Cost:1Model Size:9BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Sep 19, 2026License:apache-2.0Architecture:Transformer0.0K Open Weights Featherless Exclusive Cold

Nexus-9B-CodeCore-Merge by prithivMLmods is a 9-billion parameter merged language model based on Qwen3.5-9B, designed for advanced coding and reasoning tasks. It integrates capabilities from NeoHorse-1-9B, OmniCoder-9B, and Ornith-1.5-9B to excel in long-horizon coding, agentic workflows, and multi-step problem solving. This model is optimized for complex software engineering tasks, including code generation, understanding, modification, and debugging, with a context length of 32768 tokens.

Loading preview...

Nexus-9B-CodeCore-Merge Overview

Nexus-9B-CodeCore-Merge is a 9-billion parameter model developed by prithivMLmods, specifically engineered for sophisticated coding and reasoning applications. Built upon the Qwen3.5-9B base, it incorporates specialized capabilities from NeoHorse-1-9B, OmniCoder-9B, and Ornith-1.5-9B through a strategic merge. This fusion aims to create a robust model for handling complex software development challenges.

Key Capabilities

  • Long-Horizon Coding: Designed to manage and execute coding tasks that span multiple steps and require sustained logical progression.
  • Agentic Coding & Reasoning: Optimized for autonomous coding workflows, enabling the model to act as an intelligent agent in development environments.
  • Multi-Step Problem Solving: Excels at breaking down and solving intricate problems that demand sequential reasoning and decision-making.
  • Code Understanding & Modification: Capable of interpreting existing codebases, identifying issues, and implementing necessary changes or debugging.
  • Instruction Following: Enhanced ability to adhere to complex instructions for generating or manipulating code.

Good For

  • Complex Software Engineering: Ideal for tasks requiring deep understanding and manipulation of code.
  • Automated Development Workflows: Suitable for integrating into agent-based systems for code generation, review, and debugging.
  • Research in AI Agents: Provides a strong foundation for experimenting with agentic reasoning in coding contexts.

This model is experimental and focuses on tool-oriented agentic workflows, making it a specialized choice for developers and researchers pushing the boundaries of AI-assisted software development.

Finetunes
2 models