What are the hardware requirements for Qwen3.8 27B Heretic NVFP4 DGX Spark?

Qwen3.8 27B Heretic NVFP4 DGX Spark requires approximately 22 GB of VRAM. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Qwen3.8 27B Heretic NVFP4 DGX Spark and how large is it?

Qwen3.8 27B Heretic NVFP4 DGX Spark is developed by SmokeyPete, a model with 27b 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 27B Heretic NVFP4 DGX Spark?

Qwen3.8 27B Heretic NVFP4 DGX Spark is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Qwen3.8 27B Heretic NVFP4 DGX Spark by SmokeyPete — VRAM 22GB

Name: Qwen3.8 27B Heretic NVFP4 DGX Spark
Author: SmokeyPete

Qwen3.8 27B Heretic NVFP4 DGX Spark is an open-source language model by SmokeyPete. Features: 27b LLM, VRAM: 22GB, License: apache-2.0, LLM Explorer Score: 0.29.

8-bit Abliterated Base model:quantized:qwen/qwen... Base model:qwen/qwen3.8-27b Conversational Dgx-spark Fp4 Gb10 Heretic Image-text-to-text Modelopt Nvfp4 Quantized Qwen3.8 Qwen3 5 Region:us Safetensors Sglang Sharded Tensorflow Uncensored

Model Card on HF 🤗: https://huggingface.co/SmokeyPete/Qwen3.8-27B-Heretic-NVFP4-DGX-Spark

Qwen3.8 27B Heretic NVFP4 DGX Spark Parameters and Internals

LLM Name	Qwen3.8 27B Heretic NVFP4 DGX Spark
Repository 🤗	https://huggingface.co/SmokeyPete/Qwen3.8-27B-Heretic-NVFP4-DGX-Spark
Base Model(s)	Qwen/Qwen3.8-27B Qwen/Qwen3.8-27B
Model Size	27b
Required VRAM	22 GB
Updated	2026-10-01
Maintainer	SmokeyPete
Model Type	qwen3_5
Model Files	10.0 GB: 1-of-3 10.0 GB: 2-of-3 2.0 GB: 3-of-3
Model Architecture	Qwen3_5ForConditionalGeneration
License	apache-2.0
Model Max Length	262144
Transformers Version	5.12.1
Tokenizer Class	Qwen2Tokenizer
Padding Token	<\|im_end\|>
Errors	replace

Rank the Qwen3.8 27B Heretic NVFP4 DGX Spark 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

Best Alternatives to Qwen3.8 27B Heretic NVFP4 DGX Spark

Best Alternatives	Context / RAM	Downloads	Likes
...wopus3.8 27B Flash Mlx OQ4e 1M	1024K / 16.9 GB	211	1
Agens Pilot	1024K / 54.7 GB	113	2
Qwen3.6 27B	0K / 55.4 GB	6728138	2197
Qwen3.8 27B	0K / 55.6 GB	2	8753
Qwen3.8 27B	0K / 55.6 GB	2727	21
Qwen3.8 27B FP8	0K / 0.4 GB	123157	416
Qwen3.8 27B MLX	0K / 16 GB	1411	18
Qwen3.6 27B NVFP4	0K / 22 GB	1887525	419
Qwen3.5 27B	0K / 55.2 GB	2536294	1028
Qwen3.6 27B FP8	0K / 0.4 GB	7740472	342

Note: green Score (e.g. "73.2") means that the model is better than SmokeyPete/Qwen3.8-27B-Heretic-NVFP4-DGX-Spark.