Outmake Models
Models

Groot 19B A3B 261023

GrootTeam · October 2026

Get Hugging Face

A 19B, 3.5B active chat model from GrootTeam that reads pictures as well as text.

Downloads
28K
Likes
62
Parameters
19B, 3.5B active
Context
256K
Licence
apache-2.0

Runs well on M4 · 16 GB

The Q3_K_L size from bartowski is the one to get: a 10 GB download that uses 11 GB of memory, and writes about 16 words a second.

Groot 19B A3B 261023 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 bartowski/GrootTeam_Groot-19B-A3B-261023-GGUF on Hugging Face. A smaller size is a smaller download and a little less accurate.

SizeDownloadMemory at 8KGet
IQ2_S6.7 GB6.9 GBestimatedOpen
IQ2_M7.3 GB7.5 GBestimatedOpen
Q2_K7.9 GB8.1 GBestimatedOpen
IQ3_XXS8.8 GB9.0 GBestimatedOpen
Q3_K_S9.0 GB9.2 GBestimatedOpen
IQ3_XS9.0 GB9.2 GBestimatedOpen
Q3_K_M9.6 GB9.8 GBestimatedOpen
Q3_K_L10 GB10 GBestimatedOpen
IQ3_M11 GB11 GBestimatedOpen
IQ4_XS11 GB11 GBestimatedOpen
Q4_011 GB11 GBestimatedOpen
Q4_K_S12 GB12 GBestimatedOpen
Q4_112 GB12 GBestimatedOpen
IQ4_NL12 GB13 GBestimatedOpen
Q4_K_L13 GB14 GBestimatedOpen
Q5_K_S14 GB14 GBestimatedOpen
Q5_K_M15 GB15 GBestimatedOpen
Q6_K_S16 GB16 GBestimatedOpen
Q6_K17 GB17 GBestimatedOpen
Q6_K_L18 GB18 GBestimatedOpen
Q8_020 GB20 GBestimatedOpen
BF1638 GB38 GBestimatedOpen

Run It Anywhere

Groot 19B A3B 261023 is an open model, so any app that runs GGUF files can run it. These take the Q3_K_L size from bartowski straight from Hugging Face.

  • LM StudioMac, Windows, LinuxOpen
  • JanMac, Windows, LinuxOpen
  • MstyMac, Windows, LinuxOpen
  • OllamaMac, Windows, Linuxollama run hf.co/bartowski/GrootTeam_Groot-19B-A3B-261023-GGUF:Q3_K_L
  • llama.cppMac, Windows, Linuxllama-server -hf bartowski/GrootTeam_Groot-19B-A3B-261023-GGUF:Q3_K_L
  • Docker Model RunnerMac, Windows, Linuxdocker model run hf.co/bartowski/GrootTeam_Groot-19B-A3B-261023-GGUF:Q3_K_L
On a PC Graphics Card
Video memoryLargest size that fitsDownloadMemory at 8K
6 GBNone worth runningneeds 9.0 GB
8 GBNone worth runningneeds 9.0 GB
10 GBQ3_K_M9.6 GB9.8 GB
11 GBQ3_K_L10 GB10 GB
12 GBQ4_K_S12 GB12 GB
16 GBQ5_K_M15 GB15 GB
20 GBQ6_K_L18 GB18 GB
24 GBQ8_020 GB20 GB
32 GBQ8_020 GB20 GB
48 GBBF1638 GB38 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
MacSizeFitMemoryWrites
M6 · 16 GBQ3_K_LRuns well11 GB24
M6 · 24 GBQ4_0Runs great12 GB26
M6 · 32 GBQ5_K_MRuns great15 GB18
M5 · 16 GBQ3_K_LRuns well11 GB23
M5 · 24 GBQ4_0Runs great12 GB24
M5 · 32 GBQ5_K_MRuns great15 GB17
M5 Pro · 24 GBQ4_0Runs great12 GB45
M5 Pro · 48 GBQ8_0Runs great22 GB30
M5 Pro · 64 GBQ8_0Runs great23 GB30
M5 Max · 36 GBQ6_KRuns great17 GB44
M5 Max · 48 GBQ8_0Runs great22 GB55
M5 Max · 64 GBQ8_0Runs great23 GB55
M5 Max · 128 GBQ8_0Runs great23 GB55
M5 Ultra · 96 GBQ8_0Runs great23 GB94
M5 Ultra · 256 GBQ8_0Runs great23 GB94
M5 Ultra · 512 GBQ8_0Runs great23 GB94
M4 · 16 GBQ3_K_LRuns well11 GB16
M4 · 24 GBQ4_0Runs great12 GB17
M4 · 32 GBQ5_K_MRuns great15 GB12
M4 Pro · 24 GBQ4_0Runs great12 GB30
M4 Pro · 48 GBQ8_0Runs great22 GB22
M4 Pro · 64 GBQ8_0Runs great23 GB22
M4 Max · 36 GBQ6_KRuns great17 GB28
M4 Max · 48 GBQ8_0Runs great22 GB34
M4 Max · 64 GBQ8_0Runs great23 GB34
M4 Max · 96 GBQ8_0Runs great23 GB29
M4 Max · 128 GBQ8_0Runs great23 GB34
M3 · 8 GBIQ2_SToo big9.9 GB
M3 · 16 GBQ3_K_LRuns well11 GB14
M3 · 24 GBQ4_0Runs great12 GB15
M3 Pro · 18 GBQ4_K_MRuns well12 GB18
M3 Pro · 36 GBQ6_KRuns great17 GB14
M3 Max · 36 GBQ6_KRuns great17 GB24
M3 Max · 48 GBQ8_0Runs great22 GB30
M3 Max · 64 GBQ8_0Runs great23 GB30
M3 Max · 96 GBQ8_0Runs great23 GB24
M3 Max · 128 GBQ8_0Runs great23 GB30
M3 Ultra · 96 GBQ8_0Runs great23 GB46
M3 Ultra · 256 GBQ8_0Runs great23 GB46
M3 Ultra · 512 GBQ8_0Runs great23 GB46
M2 · 8 GBIQ2_SToo big9.9 GB
M2 · 16 GBQ3_K_LRuns well11 GB14
M2 · 24 GBQ4_0Runs great12 GB15
M2 Pro · 16 GBQ3_K_LRuns well11 GB24
M2 Pro · 32 GBQ5_K_MRuns great15 GB19
M2 Max · 32 GBQ5_K_MRuns great15 GB31
M2 Max · 64 GBQ8_0Runs great23 GB30
M2 Max · 96 GBQ8_0Runs great23 GB30
M2 Ultra · 64 GBQ8_0Runs great23 GB45
M2 Ultra · 128 GBQ8_0Runs great23 GB45
M2 Ultra · 192 GBQ8_0Runs great23 GB45
M1 · 8 GBIQ2_SToo big9.9 GB
M1 · 16 GBQ3_K_LRuns well11 GB9.2
M1 Pro · 16 GBQ3_K_LRuns well11 GB20
M1 Pro · 32 GBQ5_K_MRuns great15 GB17
M1 Max · 32 GBQ5_K_MRuns great15 GB25
M1 Max · 64 GBQ8_0Runs great23 GB25
M1 Ultra · 64 GBQ8_0Runs great23 GB35
M1 Ultra · 128 GBQ8_0Runs great23 GB35

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

Packaged by

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

Information

Maker
GrootTeam
Released
October 2026
Parameters
19B, 3.5B active
Longest chat
256K tokens
Reads pictures
Yes
Licence
apache-2.0
Downloads this month
28,003
Likes
62
Sizes
22, 6.7 to 38 GB
Kind
Maker's release

Badge for the model card

Shows on Groot 19B A3B 261023's Hugging Face card which Macs run it well, linked to this page. It updates every morning.

Runs on Mac badge

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.

Your Computer

Every model is judged for this Mac. Both are under About This Mac, in the Apple menu.

Every Mac · Every graphics card

Civitai

Open in OutmakeHugging Face

New to Outmake? Download the free beta