prithivMLmods/Orion-9B-Agentic-CodeCore-Merge
Orion-9B-Agentic-CodeCore-Merge is a 9-billion parameter merged language model developed by prithivMLmods, built upon Qwen3.5-9B. It integrates capabilities from MiMo-V2.6-Distill-Qwen-9B, OxCoder-9B, and Ornith-1.5-9B to specialize in long-horizon coding tasks, agentic coding, and complex software engineering. This model is optimized for multi-step problem solving, code understanding, modification, debugging, and tool-oriented agentic workflows, supporting a 32768-token context length.
Loading preview...
Orion-9B-Agentic-CodeCore-Merge: Advanced Agentic Coding Model
Orion-9B-Agentic-CodeCore-Merge is a 9-billion parameter model developed by prithivMLmods, specifically engineered for advanced coding and reasoning tasks. It is constructed by merging several specialized models, including Qwen3.5-9B as its base, along with MiMo-V2.6-Distill-Qwen-9B, OxCoder-9B, and Ornith-1.5-9B. This unique merge process, utilizing omnimergekit, combines their strengths to create a model proficient in complex, multi-step coding scenarios.
Key Capabilities
- Long-Horizon Coding: Designed to handle coding tasks that require sustained reasoning over extended periods.
- Agentic Coding & Reasoning: Excels in autonomous coding workflows, multi-step problem solving, and instruction following.
- Code Understanding & Modification: Capable of interpreting, modifying, and debugging code effectively.
- Tool-Oriented Workflows: Intended for integration into agentic systems that leverage external tools.
Good For
- Complex Software Engineering: Ideal for tasks demanding deep code comprehension and iterative development.
- Automated Development Agents: Suitable for building agents that can autonomously tackle coding challenges.
- Multi-Step Problem Solving: Addresses problems requiring a sequence of logical steps and code manipulation.
This experimental model, with its 32768-token context length, aims to push the boundaries of what LLMs can achieve in the domain of agentic software development.