Models

Grug 27B

ProCreations · July 2026

Get

Hugging Face

A 27B chat model from ProCreations that reads pictures as well as text.

Downloads: 360K
Likes: 41
Parameters: 27B
Context: 256K
Licence: apache-2.0

Runs well on M4 · 16 GB

The IQ2_S size from bartowski is the one to get: a 10 GB download that uses 10 GB of memory, and writes about 5.3 words a second.

Grug 27B runs well from a Mac with an M1 and 16 GB of memory. 56 of 59 Mac configurations run it well.

Compare with

The nearest models in size.

Sizes

From ProCreations/grug-27b-gguf on Hugging Face. A smaller size is a smaller download and a little less accurate.

Size	Download	Memory at 8K	Get
Q3_K_M	13 GB	14 GBestimated	Open
Q4_K_M	17 GB	17 GBestimated	Open
Q5_K_M	19 GB	20 GBestimated	Open
Q6_K	22 GB	23 GBestimated	Open
Q8_0	29 GB	29 GBestimated	Open

Run It Anywhere

Grug 27B is an open model, so any app that runs GGUF files can run it. These take the IQ2_S size from bartowski straight from Hugging Face.

LM StudioMac, Windows, LinuxOpen
JanMac, Windows, LinuxOpen
MstyMac, Windows, LinuxOpen
OllamaMac, Windows, Linuxollama run hf.co/bartowski/ProCreations_grug-27b-GGUF:IQ2_S
llama.cppMac, Windows, Linuxllama-server -hf bartowski/ProCreations_grug-27b-GGUF:IQ2_S
Docker Model RunnerMac, Windows, Linuxdocker model run hf.co/bartowski/ProCreations_grug-27b-GGUF:IQ2_S

On a PC Graphics Card

Video memory	Largest size that fits	Download	Memory at 8K
6 GB	None worth running		needs 14 GB
8 GB	None worth running		needs 14 GB
10 GB	None worth running		needs 14 GB
11 GB	None worth running		needs 14 GB
12 GB	None worth running		needs 14 GB
16 GB	Q3_K_M	13 GB	14 GB
20 GB	Q5_K_M	19 GB	20 GB
24 GB	Q6_K	22 GB	23 GB
32 GB	Q8_0	29 GB	29 GB
48 GB	Q8_0	29 GB	29 GB

The largest size, of those that keep three bits a weight or more, whose memory at an 8K chat is under 92% of the card's; the rest is the desktop's. Estimated from the files, the same way as for a Mac; speed on a PC is not estimated yet. A size that doesn't fit the card still runs, slower, with part of it in ordinary memory.

On every Mac

Mac	Size	Fit	Memory	Writes
M6 · 16 GB	IQ2_S	Runs well	10 GB	7.6
M6 · 24 GB	IQ4_XS	Runs well	16 GB	5.4
M6 · 32 GB	Q3_K_M	Runs great	15 GB	6
M5 · 16 GB	IQ2_S	Runs well	10 GB	6.9
M5 · 24 GB	IQ4_XS	Runs well	16 GB	4.9
M5 · 32 GB	Q3_K_M	Runs great	15 GB	5.5
M5 Pro · 24 GB	IQ4_XS	Runs well	16 GB	9.6
M5 Pro · 48 GB	Q6_K	Runs great	23 GB	7.4
M5 Pro · 64 GB	Q8_0	Runs great	30 GB	6.6
M5 Max · 36 GB	Q4_K_M	Runs great	17 GB	13
M5 Max · 48 GB	Q6_K	Runs great	23 GB	14
M5 Max · 64 GB	Q8_0	Runs great	30 GB	13
M5 Max · 128 GB	Q8_0	Runs great	39 GB	13
M5 Ultra · 96 GB	Q8_0	Runs great	39 GB	25
M5 Ultra · 256 GB	Q8_0	Runs great	39 GB	25
M5 Ultra · 512 GB	Q8_0	Runs great	39 GB	25
M4 · 16 GB	IQ2_S	Runs well	10 GB	5.3
M4 · 24 GB	IQ4_XS	Runs well	16 GB	3.8
M4 · 32 GB	Q3_K_M	Runs great	15 GB	4.3
M4 Pro · 24 GB	IQ4_XS	Runs well	16 GB	8.1
M4 Pro · 48 GB	Q6_K	Runs great	23 GB	6.1
M4 Pro · 64 GB	Q8_0	Runs great	30 GB	5.6
M4 Max · 36 GB	Q4_K_M	Runs great	17 GB	11
M4 Max · 48 GB	Q6_K	Runs great	23 GB	11
M4 Max · 64 GB	Q8_0	Runs great	30 GB	10
M4 Max · 96 GB	Q8_0	Runs great	39 GB	8
M4 Max · 128 GB	Q8_0	Runs great	39 GB	10
M3 · 8 GB	IQ2_XXS	Too big	20 GB
M3 · 16 GB	IQ2_S	Runs well	10 GB	4.6
M3 · 24 GB	IQ4_XS	Runs well	16 GB	3.2
M3 Pro · 18 GB	IQ2_M	Runs well	11 GB	6.4
M3 Pro · 36 GB	Q4_K_M	Runs great	17 GB	4.5
M3 Max · 36 GB	Q4_K_M	Runs great	17 GB	8.5
M3 Max · 48 GB	Q6_K	Runs great	23 GB	8.7
M3 Max · 64 GB	Q8_0	Runs great	30 GB	8
M3 Max · 96 GB	Q8_0	Runs great	39 GB	6.1
M3 Max · 128 GB	Q8_0	Runs great	39 GB	8
M3 Ultra · 96 GB	Q8_0	Runs great	39 GB	15
M3 Ultra · 256 GB	Q8_0	Runs great	39 GB	15
M3 Ultra · 512 GB	Q8_0	Runs great	39 GB	15
M2 · 8 GB	IQ2_XXS	Too big	20 GB
M2 · 16 GB	IQ2_S	Runs well	10 GB	4.5
M2 · 24 GB	IQ4_XS	Runs well	16 GB	3.2
M2 Pro · 16 GB	IQ2_S	Runs well	10 GB	8.6
M2 Pro · 32 GB	Q3_K_M	Runs great	15 GB	7
M2 Max · 32 GB	Q3_K_M	Runs great	15 GB	13
M2 Max · 64 GB	Q8_0	Runs great	30 GB	8
M2 Max · 96 GB	Q8_0	Runs great	39 GB	8
M2 Ultra · 64 GB	Q8_0	Runs great	30 GB	15
M2 Ultra · 128 GB	Q8_0	Runs great	39 GB	15
M2 Ultra · 192 GB	Q8_0	Runs great	39 GB	15
M1 · 8 GB	IQ2_XXS	Too big	20 GB
M1 · 16 GB	IQ2_S	Runs well	10 GB	3.1
M1 Pro · 16 GB	IQ2_S	Runs well	10 GB	8
M1 Pro · 32 GB	Q3_K_M	Runs great	15 GB	6.6
M1 Max · 32 GB	Q3_K_M	Runs great	15 GB	11
M1 Max · 64 GB	Q8_0	Runs great	30 GB	7.5
M1 Ultra · 64 GB	Q8_0	Runs great	30 GB	13
M1 Ultra · 128 GB	Q8_0	Runs great	39 GB	13

Writes is words a second, estimated from the Mac's memory speed and the model's size; how, and how wrong.

Packaged by

3 repositories on Hugging Face carry this model for llama.cpp. Outmake reads them all as one model.

Information

Maker: ProCreations
Released: July 2026
Parameters: 27B
Longest chat: 256K tokens
Reads pictures: Yes
Licence: apache-2.0
Downloads this month: 359,925
Likes: 41
Sizes: 5, 13 to 29 GB
Kind: Maker's release
Source: ProCreations/grug-27b

Badge for the model card

Shows on Grug 27B's Hugging Face card which Macs run it well, linked to this page. It updates every morning.

Files, downloads and likes read from Hugging Face on November 5, 2026 and refreshed every morning. Memory and speed are estimates until someone running Outmake measures this model on that Mac.