LLM EXPLORER 58,608 MODELS INDEXED

Grug 27B by ProCreations

By ProCreations · 881 downloads
HF Repo ↗ Copy URL

Grug 27B 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:qwen/qwen3...   Base model:qwen/qwen3.6-27b   Conversational Dataset:procreations/grug-thin... Dataset:procreations/grug-thin...   En   Endpoints compatible   Grug   Qwen3 5   Reasoning   Region:us   Safetensors   Sharded   Tensorflow   Token-efficient   Tool-use
Model Card on HF 🤗: https://huggingface.co/ProCreations/grug-27b 

Grug 27B Parameters and Internals

LLM NameGrug 27B
Repository 🤗https://huggingface.co/ProCreations/grug-27b 
Base Model(s)  Qwen/Qwen3.6-27B   Qwen/Qwen3.6-27B
Model Size27b
Required VRAM54.8 GB
Updated2026-07-25
MaintainerProCreations
Model Typeqwen3_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
Supported Languagesen
Model ArchitectureQwen3_5ForConditionalGeneration
Licenseapache-2.0
Model Max Length262144
Transformers Version5.14.1
Tokenizer ClassQwen2Tokenizer
Padding Token<|endoftext|>
Errorsreplace

Quantized Models of the Grug 27B

Model
Likes
Downloads
VRAM
Grug 27B Text Mlx 8Bit127428 GB

Best Alternatives to Grug 27B

Best Alternatives
Context / RAM
Downloads
Likes
Qwen3.8 27B FP80K / 0.4 GB123157416
Qwen3.8 27B0K / 55.6 GB28753
Qwen3.6 27B0K / 55.4 GB67281382197
Qwen3.6 27B NVFP40K / 22 GB1887525419
...n3.8 27B MTPLX Optimized Speed0K / 19.6 GB698950
ThinkingCap Qwen3.6 27B0K / 54.7 GB35947614
Qwen3.6 27B FP80K / 0.4 GB7740472342
Qwen3.5 27B0K / 55.2 GB25362941028
Qwen3.5 27B0K / 55.2 GB811516
Qwen3.8 27B NVFP40K / 22 GB3012214
Note: green Score (e.g. "73.2") means that the model is better than ProCreations/grug-27b.