Outmake Models
Models

Mistral Small 2409

Mistral AI · September 2024

Get Hugging Face

A 22B chat model from Mistral AI.

Downloads
55K
Likes
59
Parameters
22B
Context
128K
Licence
other

Runs well on M4 · 16 GB

The IQ3_XS size from bartowski is the one to get: a 9.2 GB download that uses 10 GB of memory, and writes about 4.1 words a second.

Ratings

KnowledgeMMLU-Pro
48.4%

MMLU-Pro: graduate-level questions in 14 subjects. Scores are the ones filed on the model's Hugging Face page or the board's own table, read 2026-11-05; some come from the maker.

Mistral Small 2409 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/Mistral-Small-Instruct-2409-GGUF on Hugging Face. A smaller size is a smaller download and a little less accurate.

SizeDownloadMemory at 8KGet
IQ2_XXS6.0 GB7.2 GBestimatedOpen
IQ2_XS6.6 GB7.8 GBestimatedOpen
IQ2_M7.6 GB8.8 GBestimatedOpen
Q2_K8.3 GB9.5 GBestimatedOpen
Q2_K_L8.5 GB9.7 GBestimatedOpen
IQ3_XS9.2 GB10 GBestimatedOpen
Q3_K_S9.6 GB11 GBestimatedOpen
IQ3_M10 GB11 GBestimatedOpen
Q3_K_M11 GB12 GBestimatedOpen
Q3_K_L12 GB13 GBestimatedOpen
Q3_K_XL12 GB13 GBestimatedOpen
IQ4_XS12 GB13 GBestimatedOpen
Q4_0_4_413 GB14 GBestimatedOpen
Q4_0_4_813 GB14 GBestimatedOpen
Q4_0_8_813 GB14 GBestimatedOpen
Q4_013 GB14 GBestimatedOpen
Q4_K_S13 GB14 GBestimatedOpen
Q4_K_M13 GB15 GBestimatedOpen
Q4_K_L13 GB15 GBestimatedOpen
Q5_K_S15 GB17 GBestimatedOpen
Q5_K_M16 GB17 GBestimatedOpen
Q5_K_L16 GB17 GBestimatedOpen
Q6_K18 GB19 GBestimatedOpen
Q6_K_L18 GB20 GBestimatedOpen
Q8_024 GB25 GBestimatedOpen
F1644 GB46 GBestimatedOpen

Run It Anywhere

Mistral Small 2409 is an open model, so any app that runs GGUF files can run it. These take the IQ3_XS size from bartowski straight from Hugging Face.

  • LM StudioMac, Windows, LinuxOpen
  • JanMac, Windows, LinuxOpen
  • MstyMac, Windows, LinuxOpen
  • OllamaMac, Windows, Linuxollama run hf.co/bartowski/Mistral-Small-Instruct-2409-GGUF:IQ3_XS
  • llama.cppMac, Windows, Linuxllama-server -hf bartowski/Mistral-Small-Instruct-2409-GGUF:IQ3_XS
  • Docker Model RunnerMac, Windows, Linuxdocker model run hf.co/bartowski/Mistral-Small-Instruct-2409-GGUF:IQ3_XS
On a PC Graphics Card
Video memoryLargest size that fitsDownloadMemory at 8K
6 GBNone worth runningneeds 10 GB
8 GBNone worth runningneeds 10 GB
10 GBNone worth runningneeds 10 GB
11 GBQ3_K_S9.6 GB11 GB
12 GBIQ3_M10 GB11 GB
16 GBQ4_K_L13 GB15 GB
20 GBQ6_K_L18 GB20 GB
24 GBQ6_K_L18 GB20 GB
32 GBQ8_024 GB25 GB
48 GBF1644 GB46 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 GBIQ3_XSRuns well10 GB5.8
M6 · 24 GBQ4_K_LRuns well16 GB4.6
M6 · 32 GBQ4_K_LRuns great15 GB4.6
M5 · 16 GBIQ3_XSRuns well10 GB5.3
M5 · 24 GBQ4_K_LRuns well16 GB4.2
M5 · 32 GBQ4_K_LRuns great15 GB4.2
M5 Pro · 24 GBQ4_K_LRuns well16 GB8.2
M5 Pro · 48 GBQ6_K_LRuns great23 GB7
M5 Pro · 64 GBQ8_0Runs great28 GB6.6
M5 Max · 36 GBQ5_K_LRuns great17 GB11
M5 Max · 48 GBQ6_K_LRuns great23 GB14
M5 Max · 64 GBQ8_0Runs great28 GB13
M5 Max · 128 GBQ8_0Runs great41 GB13
M5 Ultra · 96 GBQ8_0Runs great41 GB25
M5 Ultra · 256 GBQ8_0Runs great41 GB25
M5 Ultra · 512 GBQ8_0Runs great41 GB25
M4 · 16 GBIQ3_XSRuns well10 GB4.1
M4 · 24 GBQ4_K_LRuns well16 GB3.3
M4 · 32 GBQ4_K_LRuns great15 GB3.3
M4 Pro · 24 GBQ4_K_LRuns well16 GB7
M4 Pro · 48 GBQ6_K_LRuns great23 GB5.8
M4 Pro · 64 GBQ8_0Runs great28 GB5.6
M4 Max · 36 GBQ5_K_LRuns great17 GB9.2
M4 Max · 48 GBQ6_K_LRuns great23 GB11
M4 Max · 64 GBQ8_0Runs great28 GB10
M4 Max · 96 GBQ8_0Runs great41 GB8.1
M4 Max · 128 GBQ8_0Runs great41 GB10
M3 · 8 GBIQ1_SToo big22 GB
M3 · 16 GBIQ3_XSRuns well10 GB3.5
M3 · 24 GBQ4_K_LRuns well16 GB2.8
M3 Pro · 18 GBQ3_K_MRuns well12 GB4.7
M3 Pro · 36 GBQ5_K_LRuns great17 GB3.7
M3 Max · 36 GBQ5_K_LRuns great17 GB7.1
M3 Max · 48 GBQ6_K_LRuns great23 GB8.4
M3 Max · 64 GBQ8_0Runs great28 GB8
M3 Max · 96 GBQ8_0Runs great41 GB6.2
M3 Max · 128 GBQ8_0Runs great41 GB8
M3 Ultra · 96 GBQ8_0Runs great41 GB15
M3 Ultra · 256 GBQ8_0Runs great41 GB15
M3 Ultra · 512 GBQ8_0Runs great41 GB15
M2 · 8 GBIQ1_SToo big22 GB
M2 · 16 GBIQ3_XSRuns well10 GB3.5
M2 · 24 GBQ4_K_LRuns well16 GB2.8
M2 Pro · 16 GBIQ3_XSRuns well10 GB6.6
M2 Pro · 32 GBQ4_K_LRuns great15 GB5.3
M2 Max · 32 GBQ4_K_LRuns great15 GB10
M2 Max · 64 GBQ8_0Runs great28 GB8
M2 Max · 96 GBQ8_0Runs great41 GB8
M2 Ultra · 64 GBQ8_0Runs great28 GB15
M2 Ultra · 128 GBQ8_0Runs great41 GB15
M2 Ultra · 192 GBQ8_0Runs great41 GB15
M1 · 8 GBIQ1_SToo big22 GB
M1 · 16 GBIQ3_XSRuns well10 GB2.3
M1 Pro · 16 GBIQ3_XSRuns well10 GB6.3
M1 Pro · 32 GBQ4_K_LRuns great15 GB5.1
M1 Max · 32 GBQ4_K_LRuns great15 GB9.4
M1 Max · 64 GBQ8_0Runs great28 GB7.6
M1 Ultra · 64 GBQ8_0Runs great28 GB14
M1 Ultra · 128 GBQ8_0Runs great41 GB14

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
Mistral AI
Released
September 2024
Parameters
22B
Longest chat
128K tokens
Reads pictures
No
Licence
other
Downloads this month
55,284
Likes
59
Sizes
26, 6.0 to 44 GB
Kind
Maker's release

Badge for the model card

Shows on Mistral Small 2409'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