Outmake Models
Models

SmolLM3 3B

Hugging Face · July 2025

Get Hugging Face

A 3.1B chat model from Hugging Face.

Downloads
93K
Likes
128
Parameters
3.1B
Context
64K
Licence
apache-2.0

Runs great on M4 · 16 GB

The Q8_0 size from unsloth is the one to get: a 3.3 GB download that uses 6.2 GB of memory, and writes about 13 words a second.

SmolLM3 3B 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 unsloth/SmolLM3-3B-GGUF on Hugging Face. A smaller size is a smaller download and a little less accurate.

SizeDownloadMemory at 8KGet
UD-IQ2_XXS1.0 GB1.4 GBestimatedOpen
UD-IQ2_M1.2 GB1.6 GBestimatedOpen
Q2_K1.3 GB1.7 GBestimatedOpen
Q2_K_L1.3 GB1.7 GBestimatedOpen
UD-IQ3_XXS1.3 GB1.7 GBestimatedOpen
UD-Q2_K_XL1.3 GB1.7 GBestimatedOpen
Q3_K_S1.4 GB1.9 GBestimatedOpen
Q3_K_M1.6 GB2.0 GBestimatedOpen
UD-Q3_K_XL1.6 GB2.1 GBestimatedOpen
IQ4_XS1.7 GB2.2 GBestimatedOpen
IQ4_NL1.8 GB2.3 GBestimatedOpen
Q4_01.8 GB2.3 GBestimatedOpen
Q4_K_S1.8 GB2.3 GBestimatedOpen
Q4_K_M1.9 GB2.4 GBestimatedOpen
UD-Q4_K_XL1.9 GB2.4 GBestimatedOpen
Q4_12.0 GB2.4 GBestimatedOpen
Q5_K_S2.2 GB2.6 GBestimatedOpen
Q5_K_M2.2 GB2.7 GBestimatedOpen
UD-Q5_K_XL2.2 GB2.7 GBestimatedOpen
Q6_K2.5 GB3.0 GBestimatedOpen
UD-Q6_K_XL2.8 GB3.2 GBestimatedOpen
Q8_03.3 GB3.7 GBestimatedOpen
UD-Q8_K_XL3.9 GB4.4 GBestimatedOpen
BF166.2 GB6.6 GBestimatedOpen

Run It Anywhere

SmolLM3 3B is an open model, so any app that runs GGUF files can run it. These take the Q8_0 size from unsloth straight from Hugging Face.

  • LM StudioMac, Windows, LinuxOpen
  • JanMac, Windows, LinuxOpen
  • MstyMac, Windows, LinuxOpen
  • OllamaMac, Windows, Linuxollama run hf.co/unsloth/SmolLM3-3B-GGUF:Q8_0
  • llama.cppMac, Windows, Linuxllama-server -hf unsloth/SmolLM3-3B-GGUF:Q8_0
  • Docker Model RunnerMac, Windows, Linuxdocker model run hf.co/unsloth/SmolLM3-3B-GGUF:Q8_0
On a PC Graphics Card
Video memoryLargest size that fitsDownloadMemory at 8K
6 GBUD-Q8_K_XL3.9 GB4.4 GB
8 GBBF166.2 GB6.6 GB
10 GBBF166.2 GB6.6 GB
11 GBBF166.2 GB6.6 GB
12 GBBF166.2 GB6.6 GB
16 GBBF166.2 GB6.6 GB
20 GBBF166.2 GB6.6 GB
24 GBBF166.2 GB6.6 GB
32 GBBF166.2 GB6.6 GB
48 GBBF166.2 GB6.6 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 GBQ8_0Runs great6.2 GB19
M6 · 24 GBQ8_0Runs great6.2 GB19
M6 · 32 GBQ8_0Runs great6.2 GB19
M5 · 16 GBQ8_0Runs great6.2 GB18
M5 · 24 GBQ8_0Runs great6.2 GB18
M5 · 32 GBQ8_0Runs great6.2 GB18
M5 Pro · 24 GBQ8_0Runs great6.2 GB35
M5 Pro · 48 GBQ8_0Runs great6.2 GB35
M5 Pro · 64 GBQ8_0Runs great6.2 GB35
M5 Max · 36 GBQ8_0Runs great6.2 GB50
M5 Max · 48 GBQ8_0Runs great6.2 GB65
M5 Max · 64 GBQ8_0Runs great6.2 GB65
M5 Max · 128 GBQ8_0Runs great6.2 GB65
M5 Ultra · 96 GBQ8_0Runs great6.2 GB115
M5 Ultra · 256 GBQ8_0Runs great6.2 GB115
M5 Ultra · 512 GBQ8_0Runs great6.2 GB115
M4 · 16 GBQ8_0Runs great6.2 GB13
M4 · 24 GBQ8_0Runs great6.2 GB13
M4 · 32 GBQ8_0Runs great6.2 GB13
M4 Pro · 24 GBQ8_0Runs great6.2 GB26
M4 Pro · 48 GBQ8_0Runs great6.2 GB26
M4 Pro · 64 GBQ8_0Runs great6.2 GB26
M4 Max · 36 GBQ8_0Runs great6.2 GB36
M4 Max · 48 GBQ8_0Runs great6.2 GB44
M4 Max · 64 GBQ8_0Runs great6.2 GB44
M4 Max · 96 GBQ8_0Runs great6.2 GB36
M4 Max · 128 GBQ8_0Runs great6.2 GB44
M3 · 8 GBQ8_0Runs great3.7 GB11
M3 · 16 GBQ8_0Runs great6.2 GB11
M3 · 24 GBQ8_0Runs great6.2 GB11
M3 Pro · 18 GBQ8_0Runs great6.2 GB16
M3 Pro · 36 GBQ8_0Runs great6.2 GB16
M3 Max · 36 GBQ8_0Runs great6.2 GB29
M3 Max · 48 GBQ8_0Runs great6.2 GB37
M3 Max · 64 GBQ8_0Runs great6.2 GB37
M3 Max · 96 GBQ8_0Runs great6.2 GB29
M3 Max · 128 GBQ8_0Runs great6.2 GB37
M3 Ultra · 96 GBQ8_0Runs great6.2 GB60
M3 Ultra · 256 GBQ8_0Runs great6.2 GB60
M3 Ultra · 512 GBQ8_0Runs great6.2 GB60
M2 · 8 GBQ8_0Runs great3.7 GB11
M2 · 16 GBQ8_0Runs great6.2 GB11
M2 · 24 GBQ8_0Runs great6.2 GB11
M2 Pro · 16 GBQ8_0Runs great6.2 GB21
M2 Pro · 32 GBQ8_0Runs great6.2 GB21
M2 Max · 32 GBQ8_0Runs great6.2 GB37
M2 Max · 64 GBQ8_0Runs great6.2 GB37
M2 Max · 96 GBQ8_0Runs great6.2 GB37
M2 Ultra · 64 GBQ8_0Runs great6.2 GB59
M2 Ultra · 128 GBQ8_0Runs great6.2 GB59
M2 Ultra · 192 GBQ8_0Runs great6.2 GB59
M1 · 8 GBQ8_0Runs great3.7 GB7.5
M1 · 16 GBQ8_0Runs great6.2 GB7.5
M1 Pro · 16 GBQ8_0Runs great6.2 GB19
M1 Pro · 32 GBQ8_0Runs great6.2 GB19
M1 Max · 32 GBQ8_0Runs great6.2 GB32
M1 Max · 64 GBQ8_0Runs great6.2 GB32
M1 Ultra · 64 GBQ8_0Runs great6.2 GB47
M1 Ultra · 128 GBQ8_0Runs great6.2 GB47

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
Hugging Face
Released
July 2025
Parameters
3.1B
Longest chat
64K tokens
Reads pictures
No
Licence
apache-2.0
Downloads this month
93,151
Likes
128
Sizes
24, 1.0 to 6.2 GB
Kind
Maker's release

Badge for the model card

Shows on SmolLM3 3B'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