What are the hardware requirements for Qwen3.8 Whittle MoE 27B MLX 4bit?

Qwen3.8 Whittle MoE 27B MLX 4bit requires approximately 15.2 GB of VRAM and supports a context window of 256K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Qwen3.8 Whittle MoE 27B MLX 4bit and how large is it?

Qwen3.8 Whittle MoE 27B MLX 4bit is developed by barozp, a model with 4.2b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

Where can I download or evaluate Qwen3.8 Whittle MoE 27B MLX 4bit?

Qwen3.8 Whittle MoE 27B MLX 4bit is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Qwen3.8 Whittle MoE 27B MLX 4bit by barozp — VRAM 15.2GB, 256K context

Name: Qwen3.8 Whittle MoE 27B MLX 4bit
Author: barozp

Qwen3.8 Whittle MoE 27B MLX 4bit is an open-source language model by barozp. Features: 4.2b LLM, VRAM: 15.2GB, Context: 256K, License: apache-2.0, MoE, Quantized, LLM Explorer Score: 0.33.

4-bit 4bit Base model:logic65/qwen3.8-whi... Base model:quantized:logic65/q... Conversational Expert-pruning Mlx Mlx-lm Moe Opus-distill Quantized Qwen Qwen3 Qwen3 5 moe Reasoning Region:us Router-healing Safetensors Sharded Tensorflow Whittle

Model Card on HF 🤗: https://huggingface.co/barozp/Qwen3.8-Whittle-MoE-27B-MLX-4bit

Qwen3.8 Whittle MoE 27B MLX 4bit Parameters and Internals

LLM Name	Qwen3.8 Whittle MoE 27B MLX 4bit
Repository 🤗	https://huggingface.co/barozp/Qwen3.8-Whittle-MoE-27B-MLX-4bit
Base Model(s)	logic65/Qwen3.8-Whittle-MoE-27B-A17.8B logic65/Qwen3.8-Whittle-MoE-27B-A17.8B
Model Size	4.2b
Required VRAM	15.2 GB
Updated	2026-08-27
Maintainer	barozp
Model Type	qwen3_5_moe
Model Files	5.4 GB: 1-of-3 5.4 GB: 2-of-3 4.4 GB: 3-of-3
Quantization Type	4bit
Model Architecture	Qwen3_5MoeForCausalLM
License	apache-2.0
Context Length	262144
Model Max Length	262144
Tokenizer Class	Qwen2Tokenizer
Padding Token	<\|endoftext\|>
Vocabulary Size	248320
Errors	replace

Rank the Qwen3.8 Whittle MoE 27B MLX 4bit capabilities

Have you tried this model? Rate its performance — this feedback helps the ML community find the right model for their needs.

Instruction Following and Task Automation
Factuality and Completeness of Knowledge
Censorship and Alignment
Data Analysis and Insight Generation
Text Generation
Text Summarization and Feature Extraction
Code Generation
Multi-Language Support and Translation