LLM EXPLORER 59,358 MODELS INDEXED

Qwen3.6 27B A3B CoderX by ManniX-ITA

By ManniX-ITA · 213 downloads
HF Repo ↗ Copy URL

Qwen3.6 27B A3B CoderX is an open-source language model by ManniX-ITA. Features: 25.6b LLM, VRAM: 51.3GB, Context: 256K, License: apache-2.0, MoE, LLM Explorer Score: 0.31.

Base model:finetune:qwen/qwen3... Base model:qwen/qwen3.6-35b-a3...   Code   Conversational   Endpoints compatible   Expert-pruning   Moe   Mtp   Omnimergekit   Qwen3 5 moe text   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/ManniX-ITA/Qwen3.6-27B-A3B-CoderX 

Qwen3.6 27B A3B CoderX Parameters and Internals

LLM NameQwen3.6 27B A3B CoderX
Repository 🤗https://huggingface.co/ManniX-ITA/Qwen3.6-27B-A3B-CoderX 
Base Model(s)  Qwen/Qwen3.6-35B-A3B   Qwen/Qwen3.6-35B-A3B
Model Size25.6b
Required VRAM51.3 GB
Updated2026-08-23
MaintainerManniX-ITA
Model Typeqwen3_5_moe_text
Model Files  3.7 GB: 1-of-14   3.7 GB: 2-of-14   3.6 GB: 3-of-14   3.8 GB: 4-of-14   3.7 GB: 5-of-14   3.7 GB: 6-of-14   3.6 GB: 7-of-14   3.8 GB: 8-of-14   3.7 GB: 9-of-14   3.7 GB: 10-of-14   3.6 GB: 11-of-14   3.8 GB: 12-of-14   3.7 GB: 13-of-14   3.2 GB: 14-of-14   1.2 GB
Model ArchitectureQwen3_5MoeForCausalLM
Licenseapache-2.0
Context Length262144
Model Max Length262144
Transformers Version5.5.0
Tokenizer ClassTokenizersBackend
Padding Token<|endoftext|>
Vocabulary Size248320
Errorsreplace