prithivMLmods/Orion-9B-Agentic-CodeCore-Merge

1---2base_model:3- Qwen/Qwen3.5-9B4- XiaomiMiMo/MiMo-V2.6-Distill-Qwen-9B5- OrionLLM/OxCoder-9B6- ornith-ai/Ornith-1.5-9B7tags:8- text-generation-inference9- pytorch10- omnimergekit11- merge12- qwen3_513- reasoning14- chain-of-thought15- sft16- agent17- tool-use18- function-calling19- coder20library_name: transformers21license: apache-2.022language:23- en24pipeline_tag: text-generation25---26 27# **Orion-9B-Agentic-CodeCore-Merge**28 29> Orion-9B-Agentic-CodeCore-Merge is a merged 9B-parameter coding and reasoning model designed for long-horizon coding tasks, agentic coding, and agentic reasoning. It is built by merging [Qwen3.5-9B](https://huggingface.co/Qwen/Qwen3.5-9B) as the base model with [MiMo-V2.6-Distill-Qwen-9B](https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Distill-Qwen-9B), [OxCoder-9B](https://huggingface.co/OrionLLM/OxCoder-9B), and [Ornith-1.5-9B](https://huggingface.co/ornith-ai/Ornith-1.5-9B), combining their capabilities for code generation, multi-step problem solving, instruction following, and autonomous coding workflows. The model is intended for complex software-engineering tasks that require sustained reasoning across multiple steps, code understanding, modification, debugging, and tool-oriented agentic workflows. This model is experimental and may generate artifacts.30 31* *GGUF: [Orion-9B-Agentic-CodeCore-Merge-GGUF](https://huggingface.co/prithivMLmods/Orion-9B-Agentic-CodeCore-Merge-GGUF). Note: The Multi-Token Prediction (MTP) heads are not preserved in this format. The model runs as a standard single-token-per-step autoregressive decoder.*32 33## Merge Recipe34 35```bash36hf download XiaomiMiMo/MiMo-V2.6-Distill-Qwen-9B \37--local-dir ./models/MiMo-V2.6-Distill-Qwen-9B38 39hf download ornith-ai/Ornith-1.5-9B \40--local-dir ./models/Ornith-1.5-9B41 42hf download OrionLLM/OxCoder-9B \43--local-dir ./models/OxCoder-9B44 45hf download Qwen/Qwen3.5-9B \46--local-dir ./models/Qwen3.5-9B47```48 49```bash50python omnimergekit.py \51--base      models/Qwen3.5-9B --task-base ./models/Qwen3.5-9B \52--source    models/MiMo-V2.6-Distill-Qwen-9B \53--source    models/OxCoder-9B \54--source    models/Ornith-1.5-9B \55--weights   0.55,0.30,0.15 \56--method omnimerge_v2 --density 0.53 --darex-q 0.75 --seed 42 \57--no-auto-mlp-skip --skip-patterns visual.,mtp. \58--output prithivMLmods/Orion-9B-Agentic-CodeCore-Merge59```60 61## omnimergekit62 63Tools for model merging, expert pruning, differential competence-map extraction, and GGUF quantization — [omnimergekit](https://github.com/mann1x/omnimergekit)