Outmake Models
Models

Ministral 3 14B 2512

Mistral AI · October 2025

Get Hugging Face

A 14B chat model from Mistral AI that reads pictures as well as text.

Downloads
74K
Likes
190
Parameters
14B
Context
256K
Licence
apache-2.0

Runs great on M4 · 16 GB

The Q3_K_M size from bartowski is the one to get: a 6.7 GB download that uses 7.5 GB of memory, and writes about 5.7 words a second.

Ministral 3 14B 2512 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/mistralai_Ministral-3-14B-Instruct-2512-GGUF on Hugging Face. A smaller size is a smaller download and a little less accurate.

SizeDownloadMemory at 8KGet
IQ2_S4.5 GB5.4 GBestimatedOpen
IQ2_M4.8 GB5.7 GBestimatedOpen
Q2_K5.2 GB6.1 GBestimatedOpen
IQ3_XXS5.4 GB6.3 GBestimatedOpen
IQ3_XS5.8 GB6.7 GBestimatedOpen
Q2_K_L5.9 GB6.8 GBestimatedOpen
Q3_K_S6.1 GB6.9 GBestimatedOpen
IQ3_M6.3 GB7.1 GBestimatedOpen
Q3_K_M6.7 GB7.5 GBestimatedOpen
Q3_K_L7.2 GB8.1 GBestimatedOpen
IQ4_XS7.4 GB8.3 GBestimatedOpen
Q3_K_XL7.8 GB8.7 GBestimatedOpen
IQ4_NL7.8 GB8.7 GBestimatedOpen
Q4_07.8 GB8.7 GBestimatedOpen
Q4_K_S7.8 GB8.7 GBestimatedOpen
Q4_18.6 GB9.4 GBestimatedOpen
Q4_K_L8.7 GB9.6 GBestimatedOpen
Q5_K_S9.4 GB10 GBestimatedOpen
Q5_K_M9.6 GB10 GBestimatedOpen
Q5_K_L10 GB11 GBestimatedOpen
Q6_K_L11 GB12 GBestimatedOpen
BF1627 GB28 GBestimatedOpen

Run It Anywhere

Ministral 3 14B 2512 is an open model, so any app that runs GGUF files can run it. These take the Q3_K_M size from bartowski straight from Hugging Face.

  • LM StudioMac, Windows, LinuxOpen
  • JanMac, Windows, LinuxOpen
  • MstyMac, Windows, LinuxOpen
  • OllamaMac, Windows, Linuxollama run hf.co/bartowski/mistralai_Ministral-3-14B-Instruct-2512-GGUF:Q3_K_M
  • llama.cppMac, Windows, Linuxllama-server -hf bartowski/mistralai_Ministral-3-14B-Instruct-2512-GGUF:Q3_K_M
  • Docker Model RunnerMac, Windows, Linuxdocker model run hf.co/bartowski/mistralai_Ministral-3-14B-Instruct-2512-GGUF:Q3_K_M
On a PC Graphics Card
Video memoryLargest size that fitsDownloadMemory at 8K
6 GBNone worth runningneeds 6.3 GB
8 GBQ3_K_M6.7 GB7.5 GB
10 GBQ4_K_L8.7 GB9.6 GB
11 GBQ5_K_M9.6 GB10 GB
12 GBQ5_K_L10 GB11 GB
16 GBQ6_K_L11 GB12 GB
20 GBQ6_K_L11 GB12 GB
24 GBQ6_K_L11 GB12 GB
32 GBBF1627 GB28 GB
48 GBBF1627 GB28 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_MRuns great7.5 GB8.1
M6 · 24 GBQ5_K_LRuns great11 GB6.3
M6 · 32 GBQ8_0Runs great15 GB5.9
M5 · 16 GBQ3_K_MRuns great7.5 GB7.4
M5 · 24 GBQ5_K_LRuns great11 GB5.7
M5 · 32 GBQ8_0Runs great15 GB5.3
M5 Pro · 24 GBQ5_K_LRuns great11 GB11
M5 Pro · 48 GBQ8_0Runs great20 GB11
M5 Pro · 64 GBQ8_0Runs great26 GB11
M5 Max · 36 GBQ8_0Runs great16 GB16
M5 Max · 48 GBQ8_0Runs great20 GB21
M5 Max · 64 GBQ8_0Runs great26 GB21
M5 Max · 128 GBQ8_0Runs great38 GB21
M5 Ultra · 96 GBQ8_0Runs great38 GB39
M5 Ultra · 256 GBQ8_0Runs great38 GB39
M5 Ultra · 512 GBQ8_0Runs great38 GB39
M4 · 16 GBQ3_K_MRuns great7.5 GB5.7
M4 · 24 GBQ5_K_LRuns great11 GB4.5
M4 · 32 GBQ8_0Runs great15 GB4
M4 Pro · 24 GBQ5_K_LRuns great11 GB9.7
M4 Pro · 48 GBQ8_0Runs great20 GB8.8
M4 Pro · 64 GBQ8_0Runs great26 GB8.8
M4 Max · 36 GBQ8_0Runs great16 GB13
M4 Max · 48 GBQ8_0Runs great20 GB16
M4 Max · 64 GBQ8_0Runs great26 GB16
M4 Max · 96 GBQ8_0Runs great38 GB13
M4 Max · 128 GBQ8_0Runs great38 GB16
M3 · 8 GBUD-IQ2_XXSRuns well4.9 GB6.2
M3 · 16 GBQ3_K_MRuns great7.5 GB4.9
M3 · 24 GBQ5_K_LRuns great11 GB3.8
M3 Pro · 18 GBQ4_K_SRuns great8.7 GB6.5
M3 Pro · 36 GBQ8_0Runs great16 GB5
M3 Max · 36 GBQ8_0Runs great16 GB9.7
M3 Max · 48 GBQ8_0Runs great20 GB13
M3 Max · 64 GBQ8_0Runs great26 GB13
M3 Max · 96 GBQ8_0Runs great38 GB9.7
M3 Max · 128 GBQ8_0Runs great38 GB13
M3 Ultra · 96 GBQ8_0Runs great38 GB23
M3 Ultra · 256 GBQ8_0Runs great38 GB23
M3 Ultra · 512 GBQ8_0Runs great38 GB23
M2 · 8 GBUD-IQ2_XXSRuns well4.9 GB6.2
M2 · 16 GBQ3_K_MRuns great7.5 GB4.9
M2 · 24 GBQ5_K_LRuns great11 GB3.8
M2 Pro · 16 GBQ3_K_MRuns great7.5 GB9.3
M2 Pro · 32 GBQ8_0Runs great15 GB6.6
M2 Max · 32 GBQ8_0Runs great15 GB13
M2 Max · 64 GBQ8_0Runs great26 GB13
M2 Max · 96 GBQ8_0Runs great38 GB13
M2 Ultra · 64 GBQ8_0Runs great26 GB23
M2 Ultra · 128 GBQ8_0Runs great38 GB23
M2 Ultra · 192 GBQ8_0Runs great38 GB23
M1 · 8 GBUD-IQ2_XXSRuns well4.9 GB4.2
M1 · 16 GBQ3_K_MRuns great7.5 GB3.3
M1 Pro · 16 GBQ3_K_MRuns great7.5 GB8.9
M1 Pro · 32 GBQ8_0Runs great15 GB6.4
M1 Max · 32 GBQ8_0Runs great15 GB12
M1 Max · 64 GBQ8_0Runs great26 GB12
M1 Ultra · 64 GBQ8_0Runs great26 GB21
M1 Ultra · 128 GBQ8_0Runs great38 GB21

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

Packaged by

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

Information

Maker
Mistral AI
Released
October 2025
Parameters
14B
Longest chat
256K tokens
Reads pictures
Yes
Licence
apache-2.0
Downloads this month
74,288
Likes
190
Sizes
22, 4.5 to 27 GB
Kind
Maker's release

Badge for the model card

Shows on Ministral 3 14B 2512'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