LLM EXPLORER 63,069 MODELS INDEXED

AREX 2 27B INT4 W4A16 by numsu

By numsu · 313 downloads
HF Repo ↗ Copy URL

AREX 2 27B INT4 W4A16 is an open-source language model by numsu. Features: 27b LLM, VRAM: 18.7GB, License: apache-2.0, LLM Explorer Score: 0.31.

  Arxiv:2609.38288   Agent   Arex-2   Base model:baai/arex-2 Base model:quantized:baai/arex...   Compressed-tensors   Conversational   Image-text-to-text   Int4   Long-context   Quantized   Qwen3 5   Reasoning   Region:us   Safetensors   Sharded   Tensorflow   Tool-use   Vision   Vllm   W4a16
Model Card on HF 🤗: https://huggingface.co/numsu/AREX-2-27B-INT4-W4A16 

AREX 2 27B INT4 W4A16 Parameters and Internals

LLM NameAREX 2 27B INT4 W4A16
Repository 🤗https://huggingface.co/numsu/AREX-2-27B-INT4-W4A16 
Base Model(s)  AREX 2   BAAI/AREX-2
Model Size27b
Required VRAM18.7 GB
Updated2026-10-05
Maintainernumsu
Model Typeqwen3_5
Model Files  3.9 GB: 1-of-11   1.4 GB: 2-of-11   1.4 GB: 3-of-11   1.4 GB: 4-of-11   1.4 GB: 5-of-11   1.4 GB: 6-of-11   1.4 GB: 7-of-11   1.4 GB: 8-of-11   1.4 GB: 9-of-11   1.1 GB: 10-of-11   2.5 GB: 11-of-11
Model ArchitectureQwen3_5ForConditionalGeneration
Licenseapache-2.0
Model Max Length262144
Transformers Version5.2.0
Tokenizer ClassTokenizersBackend
Padding Token<|endoftext|>
Errorsreplace

Best Alternatives to AREX 2 27B INT4 W4A16

Best Alternatives
Context / RAM
Downloads
Likes
...wopus3.8 27B Flash Mlx OQ4e 1M1024K / 16.9 GB2111
Agens Pilot1024K / 54.7 GB982
Qwen3.8 27B NPU2256K /  GB2524
Clef0K / 54.7 GB42141253
Qwen3.6 27B0K / 55.4 GB67281382197
Qwen3.8 27B0K / 55.6 GB28753
Qwen3.8 27B0K / 55.6 GB272721
Qwen3.8 27B MLX0K / 16 GB141118
Qwen3.8 27B FP80K / 0.4 GB123157416
Qwen3.6 27B NVFP40K / 22 GB1887525419
Note: green Score (e.g. "73.2") means that the model is better than numsu/AREX-2-27B-INT4-W4A16.