Outmake Models
Models

Qwen3 32B

Alibaba Qwen · April 2025

Get Hugging Face

A 33B chat model from Alibaba Qwen.

Rating
4.0
Downloads
245K
Likes
235
Parameters
33B
Context
40K
Licence
apache-2.0

Runs well on M4 · 16 GB

The UD-IQ2_XXS size from unsloth is the one to get: a 9.3 GB download that uses 11 GB of memory, and writes about 3.8 words a second.

Ratings

4.0 out of 5, across people's blind votes on its answers and the tests below.

Text
1347
Coding
1406

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.

CodingAider polyglot
40%

Aider polyglot: coding exercises solved in six languages. 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.

Qwen3 32B 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 Qwen/Qwen3-32B-GGUF on Hugging Face. A smaller size is a smaller download and a little less accurate.

SizeDownloadMemory at 8KGet
Q4_K_M20 GB21 GBestimatedOpen
Q5_023 GB24 GBestimatedOpen
Q5_K_M23 GB25 GBestimatedOpen
Q6_K27 GB28 GBestimatedOpen
Q8_035 GB36 GBestimatedOpen

Run It Anywhere

Qwen3 32B is an open model, so any app that runs GGUF files can run it. These take the UD-IQ2_XXS size from unsloth straight from Hugging Face.

  • LM StudioMac, Windows, LinuxOpen
  • JanMac, Windows, LinuxOpen
  • MstyMac, Windows, LinuxOpen
  • OllamaMac, Windows, Linuxollama run hf.co/unsloth/Qwen3-32B-GGUF:UD-IQ2_XXS
  • llama.cppMac, Windows, Linuxllama-server -hf unsloth/Qwen3-32B-GGUF:UD-IQ2_XXS
  • Docker Model RunnerMac, Windows, Linuxdocker model run hf.co/unsloth/Qwen3-32B-GGUF:UD-IQ2_XXS
On a PC Graphics Card
Video memoryLargest size that fitsDownloadMemory at 8K
6 GBNone worth runningneeds 21 GB
8 GBNone worth runningneeds 21 GB
10 GBNone worth runningneeds 21 GB
11 GBNone worth runningneeds 21 GB
12 GBNone worth runningneeds 21 GB
16 GBNone worth runningneeds 21 GB
20 GBNone worth runningneeds 21 GB
24 GBQ4_K_M20 GB21 GB
32 GBQ6_K27 GB28 GB
48 GBQ8_035 GB36 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 GBUD-IQ2_XXSRuns well11 GB5.4
M6 · 24 GBQ3_K_SRuns well16 GB4.2
M6 · 32 GBQ4_K_MRuns well21 GB3.4
M5 · 16 GBUD-IQ2_XXSRuns well11 GB4.9
M5 · 24 GBQ3_K_SRuns well16 GB3.8
M5 · 32 GBQ4_K_MRuns well21 GB3.1
M5 Pro · 24 GBQ3_K_SRuns well16 GB7.5
M5 Pro · 48 GBQ4_K_MRuns great22 GB6.1
M5 Pro · 64 GBUD-Q6_K_XLRuns great30 GB4.6
M5 Max · 36 GBQ3_K_MRuns great17 GB10
M5 Max · 48 GBQ4_K_MRuns great22 GB12
M5 Max · 64 GBUD-Q6_K_XLRuns great30 GB9.1
M5 Max · 128 GBQ8_0Runs great40 GB9.4
M5 Ultra · 96 GBQ8_0Runs great40 GB18
M5 Ultra · 256 GBQ8_0Runs great40 GB18
M5 Ultra · 512 GBQ8_0Runs great40 GB18
M4 · 16 GBUD-IQ2_XXSRuns well11 GB3.8
M4 · 24 GBQ3_K_SRuns well16 GB3
M4 · 32 GBQ4_K_MRuns well21 GB2.4
M4 Pro · 24 GBQ3_K_SRuns well16 GB6.4
M4 Pro · 48 GBQ4_K_MRuns great22 GB5.3
M4 Pro · 64 GBUD-Q6_K_XLRuns great30 GB4.1
M4 Max · 36 GBQ3_K_MRuns great17 GB8.7
M4 Max · 48 GBQ4_K_MRuns great22 GB9.8
M4 Max · 64 GBUD-Q6_K_XLRuns great30 GB7.7
M4 Max · 96 GBQ8_0Runs great40 GB5.9
M4 Max · 128 GBQ8_0Runs great40 GB7.7
M3 · 8 GBUD-IQ1_SToo big13 GB
M3 · 16 GBUD-IQ2_XXSRuns well11 GB3.2
M3 · 24 GBQ3_K_SRuns well16 GB2.5
M3 Pro · 18 GBIQ2_SRuns well12 GB4.4
M3 Pro · 36 GBQ3_K_MRuns great17 GB3.5
M3 Max · 36 GBQ3_K_MRuns great17 GB6.7
M3 Max · 48 GBQ4_K_MRuns great22 GB7.6
M3 Max · 64 GBUD-Q6_K_XLRuns great30 GB5.9
M3 Max · 96 GBQ8_0Runs great40 GB4.5
M3 Max · 128 GBQ8_0Runs great40 GB5.8
M3 Ultra · 96 GBQ8_0Runs great40 GB11
M3 Ultra · 256 GBQ8_0Runs great40 GB11
M3 Ultra · 512 GBQ8_0Runs great40 GB11
M2 · 8 GBUD-IQ1_SToo big13 GB
M2 · 16 GBUD-IQ2_XXSRuns well11 GB3.2
M2 · 24 GBQ3_K_SRuns well16 GB2.5
M2 Pro · 16 GBUD-IQ2_XXSRuns well11 GB6.2
M2 Pro · 32 GBQ4_K_MRuns well21 GB4
M2 Max · 32 GBQ4_K_MRuns well21 GB7.6
M2 Max · 64 GBUD-Q6_K_XLRuns great30 GB5.9
M2 Max · 96 GBQ8_0Runs great40 GB5.8
M2 Ultra · 64 GBUD-Q6_K_XLRuns great30 GB11
M2 Ultra · 128 GBQ8_0Runs great40 GB11
M2 Ultra · 192 GBQ8_0Runs great40 GB11
M1 · 8 GBUD-IQ1_SToo big13 GB
M1 · 16 GBUD-IQ2_XXSRuns well11 GB2.2
M1 Pro · 16 GBUD-IQ2_XXSRuns well11 GB5.9
M1 Pro · 32 GBQ4_K_MRuns well21 GB3.9
M1 Max · 32 GBQ4_K_MRuns well21 GB7.1
M1 Max · 64 GBUD-Q6_K_XLRuns great30 GB5.6
M1 Ultra · 64 GBUD-Q6_K_XLRuns great30 GB10
M1 Ultra · 128 GBQ8_0Runs great40 GB10

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

Packaged by

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

Information

Maker
Alibaba Qwen
Released
April 2025
Parameters
33B
Longest chat
40K tokens
Reads pictures
No
Licence
apache-2.0
Downloads this month
245,242
Likes
235
Sizes
5, 20 to 35 GB
Kind
Maker's release

Badge for the model card

Shows on Qwen3 32B'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