Outmake Models
Models

Gemma 2 9B

Google · June 2024

Get Hugging Face

A 9.2B chat model from Google.

Rating
2.8
Downloads
36K
Likes
268
Parameters
9.2B
Context
8K
Licence
gemma

Runs great on M4 · 16 GB

The Q4_K_L size from bartowski is the one to get: a 6.0 GB download that uses 7.7 GB of memory, and writes about 4 words a second.

Ratings

2.8 out of 5, across people's blind votes on its answers.

Text
1267
Coding
1270

LM Arena ratings of October 28, 2026 (CC BY 4.0): people vote between two answers without knowing which model wrote which. The bar ends at 1488, the best any open model has. The Arena tests the full model; a small size on a Mac answers a little worse.

Gemma 2 9B runs well from a Mac with an M1 and 8 GB of memory. 59 of 59 Mac configurations run it well.

Compare with

The nearest models in size.

Sizes

From bartowski/gemma-2-9b-it-GGUF on Hugging Face. A smaller size is a smaller download and a little less accurate.

SizeDownloadMemory at 8KGet
IQ2_XS3.1 GB4.8 GBestimatedOpen
IQ2_S3.2 GB4.9 GBestimatedOpen
IQ2_M3.4 GB5.2 GBestimatedOpen
IQ3_XXS3.8 GB5.5 GBestimatedOpen
Q2_K3.8 GB5.5 GBestimatedOpen
Q2_K_L4.0 GB5.8 GBestimatedOpen
IQ3_XS4.1 GB5.9 GBestimatedOpen
Q3_K_S4.3 GB6.1 GBestimatedOpen
IQ3_M4.5 GB6.2 GBestimatedOpen
Q3_K_M4.8 GB6.5 GBestimatedOpen
Q3_K_L5.1 GB6.9 GBestimatedOpen
IQ4_XS5.2 GB6.9 GBestimatedOpen
Q3_K_XL5.4 GB7.1 GBestimatedOpen
Q4_K_S5.5 GB7.2 GBestimatedOpen
Q4_K_M5.8 GB7.5 GBestimatedOpen
Q4_K_L6.0 GB7.7 GBestimatedOpen
Q5_K_S6.5 GB8.2 GBestimatedOpen
Q5_K_M6.6 GB8.4 GBestimatedOpen
gemma-2-9b-it-Q4_K_M-fp166.8 GB8.6 GBestimatedOpen
Q5_K_L6.9 GB8.6 GBestimatedOpen
Q6_K7.6 GB9.3 GBestimatedOpen
Q6_K_L7.8 GB9.5 GBestimatedOpen
Q8_09.8 GB12 GBestimatedOpen
Q8_0_L11 GB12 GBestimatedOpen
F3237 GB39 GBestimatedOpen

Run It Anywhere

Gemma 2 9B is an open model, so any app that runs GGUF files can run it. These take the Q4_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/gemma-2-9b-it-GGUF:Q4_K_L
  • llama.cppMac, Windows, Linuxllama-server -hf bartowski/gemma-2-9b-it-GGUF:Q4_K_L
  • Docker Model RunnerMac, Windows, Linuxdocker model run hf.co/bartowski/gemma-2-9b-it-GGUF:Q4_K_L
On a PC Graphics Card
Video memoryLargest size that fitsDownloadMemory at 8K
6 GBIQ3_XS4.1 GB5.9 GB
8 GBQ4_K_L6.0 GB7.7 GB
10 GBQ6_K_L7.8 GB9.5 GB
11 GBQ6_K_L7.8 GB9.5 GB
12 GBQ8_09.8 GB12 GB
16 GBQ8_0_L11 GB12 GB
20 GBQ8_0_L11 GB12 GB
24 GBQ8_0_L11 GB12 GB
32 GBQ8_0_L11 GB12 GB
48 GBF3237 GB39 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 GBQ4_K_LRuns great7.7 GB5.7
M6 · 24 GBQ8_0Runs great12 GB5.4
M6 · 32 GBQ8_0Runs great12 GB5.4
M5 · 16 GBQ4_K_LRuns great7.7 GB5.2
M5 · 24 GBQ8_0Runs great12 GB4.9
M5 · 32 GBQ8_0Runs great12 GB4.9
M5 Pro · 24 GBQ8_0Runs great12 GB9.8
M5 Pro · 48 GBQ8_0Runs great12 GB9.8
M5 Pro · 64 GBQ8_0Runs great12 GB9.8
M5 Max · 36 GBQ8_0Runs great12 GB14
M5 Max · 48 GBQ8_0Runs great12 GB19
M5 Max · 64 GBQ8_0Runs great12 GB19
M5 Max · 128 GBQ8_0Runs great12 GB19
M5 Ultra · 96 GBQ8_0Runs great12 GB37
M5 Ultra · 256 GBQ8_0Runs great12 GB37
M5 Ultra · 512 GBQ8_0Runs great12 GB37
M4 · 16 GBQ4_K_LRuns great7.7 GB4
M4 · 24 GBQ8_0Runs great12 GB3.8
M4 · 32 GBQ8_0Runs great12 GB3.8
M4 Pro · 24 GBQ8_0Runs great12 GB8.2
M4 Pro · 48 GBQ8_0Runs great12 GB8.2
M4 Pro · 64 GBQ8_0Runs great12 GB8.2
M4 Max · 36 GBQ8_0Runs great12 GB12
M4 Max · 48 GBQ8_0Runs great12 GB15
M4 Max · 64 GBQ8_0Runs great12 GB15
M4 Max · 96 GBQ8_0Runs great12 GB12
M4 Max · 128 GBQ8_0Runs great12 GB15
M3 · 8 GBIQ2_MRuns well5.2 GB4
M3 · 16 GBQ4_K_LRuns great7.7 GB3.4
M3 · 24 GBQ8_0Runs great12 GB3.2
M3 Pro · 18 GBQ5_K_LRuns great8.6 GB4.8
M3 Pro · 36 GBQ8_0Runs great12 GB4.7
M3 Max · 36 GBQ8_0Runs great12 GB9.1
M3 Max · 48 GBQ8_0Runs great12 GB12
M3 Max · 64 GBQ8_0Runs great12 GB12
M3 Max · 96 GBQ8_0Runs great12 GB9.1
M3 Max · 128 GBQ8_0Runs great12 GB12
M3 Ultra · 96 GBQ8_0Runs great12 GB22
M3 Ultra · 256 GBQ8_0Runs great12 GB22
M3 Ultra · 512 GBQ8_0Runs great12 GB22
M2 · 8 GBIQ2_MRuns well5.2 GB4
M2 · 16 GBQ4_K_LRuns great7.7 GB3.4
M2 · 24 GBQ8_0Runs great12 GB3.2
M2 Pro · 16 GBQ4_K_LRuns great7.7 GB6.6
M2 Pro · 32 GBQ8_0Runs great12 GB6.2
M2 Max · 32 GBQ8_0Runs great12 GB12
M2 Max · 64 GBQ8_0Runs great12 GB12
M2 Max · 96 GBQ8_0Runs great12 GB12
M2 Ultra · 64 GBQ8_0Runs great12 GB22
M2 Ultra · 128 GBQ8_0Runs great12 GB22
M2 Ultra · 192 GBQ8_0Runs great12 GB22
M1 · 8 GBIQ2_MRuns well5.2 GB2.7
M1 · 16 GBQ4_K_LRuns great7.7 GB2.3
M1 Pro · 16 GBQ4_K_LRuns great7.7 GB6.4
M1 Pro · 32 GBQ8_0Runs great12 GB6
M1 Max · 32 GBQ8_0Runs great12 GB11
M1 Max · 64 GBQ8_0Runs great12 GB11
M1 Ultra · 64 GBQ8_0Runs great12 GB20
M1 Ultra · 128 GBQ8_0Runs great12 GB20

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
Google
Released
June 2024
Parameters
9.2B
Longest chat
8K tokens
Reads pictures
No
Licence
gemma
Downloads this month
36,013
Likes
268
Sizes
25, 3.1 to 37 GB
Kind
Maker's release

Badge for the model card

Shows on Gemma 2 9B'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