LLM EXPLORER 63,636 MODELS INDEXED

AREX 2 8bit by mlx-community

By mlx-community · 132 downloads
HF Repo ↗ Copy URL

AREX 2 8bit is an open-source language model by mlx-community. Features: 27.4b LLM, VRAM: 29.4GB, License: apache-2.0, Quantized, LLM Explorer Score: 0.3.

  Arxiv:2609.38288   8-bit   8bit   Agent   Apple-silicon   Arex-2   Base model:baai/arex-2 Base model:quantized:baai/arex...   Conversational   Deep-research   En   Image-text-to-text   Long-context   Mac   Mlx   Quantized   Qwen3.8   Qwen3 5   Reasoning   Region:us   Safetensors   Sharded   Tensorflow   Tool-use   Vision
Model Card on HF 🤗: https://huggingface.co/mlx-community/AREX-2-8bit 

AREX 2 8bit Parameters and Internals

LLM NameAREX 2 8bit
Repository 🤗https://huggingface.co/mlx-community/AREX-2-8bit 
Base Model(s)  AREX 2   BAAI/AREX-2
Model Size27.4b
Required VRAM29.4 GB
Updated2026-10-10
Maintainermlx-community
Model Typeqwen3_5
Model Files  5.3 GB: 1-of-6   5.3 GB: 2-of-6   5.3 GB: 3-of-6   5.3 GB: 4-of-6   5.3 GB: 5-of-6   2.9 GB: 6-of-6
Supported Languagesen
Quantization Type8bit
Model ArchitectureQwen3_5ForConditionalGeneration
Licenseapache-2.0
Model Max Length262144
Transformers Version5.2.0
Tokenizer ClassQwen3_5Tokenizer
Padding Token<|endoftext|>
Errorsreplace

Best Alternatives to AREX 2 8bit

Best Alternatives
Context / RAM
Downloads
Likes
AREX 2 5bit0K / 19.5 GB2312
TelecomGPT R116K / 54.7 GB36529
Openjev0K / 54.7 GB5073108
Clef OQ4e0K / 16.6 GB3099
Clef NVFP40K / 19.7 GB622
Nesi10K / 0.8 GB1092
Clef FP80K / 35.9 GB852
JAX XORTRON0K / 54.7 GB4441
AREX 2 FP80K / 35.9 GB323
Hemmingway 1 Heretic0K / 54.7 GB360
Note: green Score (e.g. "73.2") means that the model is better than mlx-community/AREX-2-8bit.