LLM EXPLORER 61,755 MODELS INDEXED

XYZ Aquila Mini OptiQ 4bit REAP 18B by mlx-community

By mlx-community · 181 downloads
HF Repo ↗ Copy URL

XYZ Aquila Mini OptiQ 4bit REAP 18B is an open-source language model by mlx-community. Features: 18b LLM, VRAM: 19.2GB, License: apache-2.0, Quantized, LLM Explorer Score: 0.28.

  Arxiv:2510.13999   4-bit   4bit   Apple-silicon Base model:mlx-community/xyz-a... Base model:quantized:mlx-commu...   Conversational   Expert-pruning   Mlx   Moe   Optiq   Quantized   Qwen3 5 moe   Reap   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/mlx-community/XYZ-Aquila-mini-OptiQ-4bit-REAP-18B 

XYZ Aquila Mini OptiQ 4bit REAP 18B Parameters and Internals

LLM NameXYZ Aquila Mini OptiQ 4bit REAP 18B
Repository 🤗https://huggingface.co/mlx-community/XYZ-Aquila-mini-OptiQ-4bit-REAP-18B 
Base Model(s)  XYZ Aquila Mini OptiQ 4bit   mlx-community/XYZ-Aquila-mini-OptiQ-4bit
Model Size18b
Required VRAM19.2 GB
Updated2026-09-20
Maintainermlx-community
Model Typeqwen3_5_moe
Model Files  4.2 GB: 1-of-5   4.2 GB: 2-of-5   4.2 GB: 3-of-5   4.0 GB: 4-of-5   2.6 GB: 5-of-5
Quantization Type4bit
Model ArchitectureQwen3_5MoeForConditionalGeneration
Licenseapache-2.0
Model Max Length262144
Transformers Version5.12.1
Tokenizer ClassQwen2Tokenizer
Padding Token<|endoftext|>
Errorsreplace

Best Alternatives to XYZ Aquila Mini OptiQ 4bit REAP 18B

Best Alternatives
Context / RAM
Downloads
Likes
...r V2.5 Dev OptiQ 4bit REAP 18B0K / 12.3 GB1313