Outmake Models
Models

Nanbeige4.3 33B

Nanbeige · October 2026

Get Hugging Face

A 33B chat model from Nanbeige.

Downloads
204K
Likes
145
Parameters
33B
Context
512K
Licence
apache-2.0

Runs well on M4 · 16 GB

The IQ1_M size from mradermacher is the one to get: a 8.3 GB download that uses 10 GB of memory, and writes about 4.5 words a second.

Nanbeige4.3 33B 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/Nanbeige_Nanbeige4.3-33B-GGUF on Hugging Face. A smaller size is a smaller download and a little less accurate.

SizeDownloadMemory at 8KGet
IQ2_XXS10 GB11 GBestimatedOpen
IQ2_XS11 GB12 GBestimatedOpen
IQ2_S11 GB12 GBestimatedOpen
IQ2_M12 GB13 GBestimatedOpen
Q2_K14 GB15 GBestimatedOpen
IQ3_XXS14 GB16 GBestimatedOpen
Q3_K_S15 GB16 GBestimatedOpen
IQ3_XS15 GB16 GBestimatedOpen
Q3_K_M16 GB17 GBestimatedOpen
Q3_K_L17 GB18 GBestimatedOpen
IQ3_M18 GB19 GBestimatedOpen
IQ4_XS19 GB20 GBestimatedOpen
Q4_019 GB20 GBestimatedOpen
Q4_K_S20 GB21 GBestimatedOpen
Q4_K_M21 GB22 GBestimatedOpen
IQ4_NL21 GB22 GBestimatedOpen
Q4_121 GB22 GBestimatedOpen
Q4_K_L22 GB23 GBestimatedOpen
Q5_K_S24 GB25 GBestimatedOpen
Q5_K_M25 GB26 GBestimatedOpen
Q6_K_S28 GB29 GBestimatedOpen
Q6_K29 GB30 GBestimatedOpen
Q6_K_L30 GB32 GBestimatedOpen
Q8_035 GB36 GBestimatedOpen
BF1666 GB67 GBestimatedOpen

Run It Anywhere

Nanbeige4.3 33B is an open model, so any app that runs GGUF files can run it. These take the IQ1_M size from mradermacher straight from Hugging Face.

  • LM StudioMac, Windows, LinuxOpen
  • JanMac, Windows, LinuxOpen
  • MstyMac, Windows, LinuxOpen
  • OllamaMac, Windows, Linuxollama run hf.co/mradermacher/Nanbeige4.3-33B-i1-GGUF:IQ1_M
  • llama.cppMac, Windows, Linuxllama-server -hf mradermacher/Nanbeige4.3-33B-i1-GGUF:IQ1_M
  • Docker Model RunnerMac, Windows, Linuxdocker model run hf.co/mradermacher/Nanbeige4.3-33B-i1-GGUF:IQ1_M
On a PC Graphics Card
Video memoryLargest size that fitsDownloadMemory at 8K
6 GBNone worth runningneeds 16 GB
8 GBNone worth runningneeds 16 GB
10 GBNone worth runningneeds 16 GB
11 GBNone worth runningneeds 16 GB
12 GBNone worth runningneeds 16 GB
16 GBIQ3_XXS14 GB16 GB
20 GBIQ4_XS19 GB20 GB
24 GBQ4_K_L22 GB23 GB
32 GBQ6_K_L30 GB32 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 GBIQ1_MRuns well10 GB6.3
M6 · 24 GBIQ3_XXSRuns well16 GB4.5
M6 · 32 GBQ4_K_SRuns well21 GB3.6
M5 · 16 GBIQ1_MRuns well10 GB5.8
M5 · 24 GBIQ3_XXSRuns well16 GB4.1
M5 · 32 GBQ4_K_SRuns well21 GB3.3
M5 Pro · 24 GBIQ3_XXSRuns well16 GB8.1
M5 Pro · 48 GBQ4_1Runs great22 GB7
M5 Pro · 64 GBQ6_KRuns great30 GB5.1
M5 Max · 36 GBIQ3_MRuns great17 GB11
M5 Max · 48 GBQ4_1Runs great22 GB14
M5 Max · 64 GBQ6_KRuns great30 GB9.9
M5 Max · 128 GBQ8_0Runs great68 GB9.7
M5 Ultra · 96 GBQ8_0Runs great43 GB19
M5 Ultra · 256 GBQ8_0Runs great100 GB19
M5 Ultra · 512 GBQ8_0Runs great100 GB19
M4 · 16 GBIQ1_MRuns well10 GB4.5
M4 · 24 GBIQ3_XXSRuns well16 GB3.2
M4 · 32 GBQ4_K_SRuns well21 GB2.6
M4 Pro · 24 GBIQ3_XXSRuns well16 GB6.9
M4 Pro · 48 GBQ4_1Runs great22 GB6.1
M4 Pro · 64 GBQ6_KRuns great30 GB4.3
M4 Max · 36 GBIQ3_MRuns great17 GB9.4
M4 Max · 48 GBQ4_1Runs great22 GB11
M4 Max · 64 GBQ6_KRuns great30 GB8.2
M4 Max · 96 GBQ8_0Runs great43 GB6.2
M4 Max · 128 GBQ8_0Runs great68 GB8
M3 · 8 GBIQ1_SToo big72 GB
M3 · 16 GBIQ1_MRuns well10 GB3.8
M3 · 24 GBIQ3_XXSRuns well16 GB2.7
M3 Pro · 18 GBIQ2_XSRuns well12 GB4.8
M3 Pro · 36 GBIQ3_MRuns great17 GB3.7
M3 Max · 36 GBIQ3_MRuns great17 GB7.2
M3 Max · 48 GBQ4_1Runs great22 GB8.8
M3 Max · 64 GBQ6_KRuns great30 GB6.2
M3 Max · 96 GBQ8_0Runs great43 GB4.6
M3 Max · 128 GBQ8_0Runs great68 GB6.1
M3 Ultra · 96 GBQ8_0Runs great43 GB12
M3 Ultra · 256 GBQ8_0Runs great100 GB12
M3 Ultra · 512 GBQ8_0Runs great100 GB12
M2 · 8 GBIQ1_SToo big72 GB
M2 · 16 GBIQ1_MRuns well10 GB3.8
M2 · 24 GBIQ3_XXSRuns well16 GB2.7
M2 Pro · 16 GBIQ1_MRuns well10 GB7.3
M2 Pro · 32 GBQ4_K_SRuns well21 GB4.3
M2 Max · 32 GBQ4_K_SRuns well21 GB8.2
M2 Max · 64 GBQ6_KRuns great30 GB6.2
M2 Max · 96 GBQ8_0Runs great43 GB6.1
M2 Ultra · 64 GBQ6_KRuns great30 GB12
M2 Ultra · 128 GBQ8_0Runs great68 GB11
M2 Ultra · 192 GBQ8_0Runs great100 GB11
M1 · 8 GBIQ1_SToo big72 GB
M1 · 16 GBIQ1_MRuns well10 GB2.6
M1 Pro · 16 GBIQ1_MRuns well10 GB6.9
M1 Pro · 32 GBQ4_K_SRuns well21 GB4.1
M1 Max · 32 GBQ4_K_SRuns well21 GB7.7
M1 Max · 64 GBQ6_KRuns great30 GB6
M1 Ultra · 64 GBQ6_KRuns great30 GB11
M1 Ultra · 128 GBQ8_0Runs great68 GB11

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
Nanbeige
Released
October 2026
Parameters
33B
Longest chat
512K tokens
Reads pictures
No
Licence
apache-2.0
Downloads this month
203,702
Likes
145
Sizes
25, 10 to 66 GB
Kind
Maker's release

Badge for the model card

Shows on Nanbeige4.3 33B'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