What are the hardware requirements for Grug V1.1 Qwen 3.8 27B Mtp?

Grug V1.1 Qwen 3.8 27B Mtp requires approximately 54.8 GB of VRAM. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Grug V1.1 Qwen 3.8 27B Mtp and how large is it?

Grug V1.1 Qwen 3.8 27B Mtp is developed by ProCreations, 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 Grug V1.1 Qwen 3.8 27B Mtp?

Grug V1.1 Qwen 3.8 27B Mtp is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Grug V1.1 Qwen 3.8 27B Mtp by ProCreations — VRAM 54.8GB

Name: Grug V1.1 Qwen 3.8 27B Mtp
Author: ProCreations

Grug V1.1 Qwen 3.8 27B Mtp is an open-source language model by ProCreations. Features: 27b LLM, VRAM: 54.8GB, License: apache-2.0, LLM Explorer Score: 0.34.

Agentic Base model:finetune:procreatio... Base model:procreations/grug-v... Conversational En Endpoints compatible Grug Image-text-to-text Mtp Qwen3 5 Region:us Safetensors Sharded Speculative-decoding Tensorflow Token-efficient

Model Card on HF 🤗: https://huggingface.co/ProCreations/grug-v1.1-qwen-3.8-27b-mtp

Grug V1.1 Qwen 3.8 27B Mtp Parameters and Internals

LLM Name	Grug V1.1 Qwen 3.8 27B Mtp
Repository 🤗	https://huggingface.co/ProCreations/grug-v1.1-qwen-3.8-27b-mtp
Base Model(s)	ProCreations/grug-v1.1-qwen-3.8-27b ProCreations/grug-v1.1-qwen-3.8-27b
Model Size	27b
Required VRAM	54.8 GB
Updated	2026-08-20
Maintainer	ProCreations
Model Type	qwen3_5
Model Files	2.5 GB: 1-of-15 3.9 GB: 2-of-15 4.0 GB: 3-of-15 4.0 GB: 4-of-15 3.9 GB: 5-of-15 4.0 GB: 6-of-15 4.0 GB: 7-of-15 3.8 GB: 8-of-15 4.0 GB: 9-of-15 4.0 GB: 10-of-15 3.8 GB: 11-of-15 4.0 GB: 12-of-15 4.0 GB: 13-of-15 4.0 GB: 14-of-15 0.9 GB: 15-of-15 0.8 GB
Supported Languages	en
Model Architecture	Qwen3_5ForConditionalGeneration
License	apache-2.0
Model Max Length	262144
Transformers Version	5.15.0
Tokenizer Class	TokenizersBackend
Padding Token	<\|endoftext\|>
Errors	replace

Rank the Grug V1.1 Qwen 3.8 27B Mtp 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 Grug V1.1 Qwen 3.8 27B Mtp

Best Alternatives	Context / RAM	Downloads	Likes
Qwen3.8 27B FP8	0K / 0.4 GB	123157	416
Qwen3.6 27B	0K / 55.4 GB	6728138	2197
Qwen3.8 27B	0K / 55.6 GB	2	8753
Qwen3.6 27B NVFP4	0K / 22 GB	1887525	419
ThinkingCap Qwen3.6 27B	0K / 54.7 GB	35947	614
Qwen3.6 27B FP8	0K / 0.4 GB	7740472	342
Qwen3.5 27B	0K / 55.2 GB	2536294	1028
Huihui Qwen3.8 27B Abliterated	0K / 55.6 GB	1485	76
Qwen3.5 27B	0K / 55.2 GB	8115	16
...n3.8 27B MTPLX Optimized Speed	0K / 19.6 GB	6989	50

Note: green Score (e.g. "73.2") means that the model is better than ProCreations/grug-v1.1-qwen-3.8-27b-mtp.