Outmake Models
Models

Seed OSS 36B

ByteDance · August 2025

Get Hugging Face

A 36B chat model from ByteDance.

Downloads
9.8K
Likes
47
Parameters
36B
Context
512K
Licence
apache-2.0

Runs well on M4 · 16 GB

The UD-IQ1_M size from unsloth is the one to get: a 9.1 GB download that uses 10 GB of memory, and writes about 3.8 words a second.

Ratings

KnowledgeMMLU-Pro
82.7%

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.

Seed OSS 36B 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 unsloth/Seed-OSS-36B-Instruct-GGUF on Hugging Face. A smaller size is a smaller download and a little less accurate.

SizeDownloadMemory at 8KGet
UD-IQ1_S8.5 GB9.8 GBestimatedOpen
UD-IQ1_M9.1 GB10 GBestimatedOpen
UD-IQ2_XXS10 GB12 GBestimatedOpen
UD-IQ2_M13 GB14 GBestimatedOpen
Q2_K14 GB15 GBestimatedOpen
Q2_K_L14 GB15 GBestimatedOpen
UD-Q2_K_XL14 GB15 GBestimatedOpen
UD-IQ3_XXS14 GB16 GBestimatedOpen
Q3_K_S16 GB17 GBestimatedOpen
Q3_K_M18 GB19 GBestimatedOpen
UD-Q3_K_XL18 GB19 GBestimatedOpen
IQ4_XS20 GB21 GBestimatedOpen
IQ4_NL21 GB22 GBestimatedOpen
Q4_021 GB22 GBestimatedOpen
Q4_K_S21 GB22 GBestimatedOpen
Q4_K_M22 GB23 GBestimatedOpen
UD-Q4_K_XL22 GB23 GBestimatedOpen
Q4_123 GB24 GBestimatedOpen
Q5_K_S25 GB26 GBestimatedOpen
Q5_K_M26 GB27 GBestimatedOpen
UD-Q5_K_XL26 GB27 GBestimatedOpen
Q6_K30 GB31 GBestimatedOpen
UD-Q6_K_XL32 GB33 GBestimatedOpen
Q8_038 GB40 GBestimatedOpen
UD-Q8_K_XL43 GB44 GBestimatedOpen
BF1672 GB74 GBestimatedOpen

Run It Anywhere

Seed OSS 36B is an open model, so any app that runs GGUF files can run it. These take the UD-IQ1_M size from unsloth straight from Hugging Face.

  • LM StudioMac, Windows, LinuxOpen
  • JanMac, Windows, LinuxOpen
  • MstyMac, Windows, LinuxOpen
  • OllamaMac, Windows, Linuxollama run hf.co/unsloth/Seed-OSS-36B-Instruct-GGUF:UD-IQ1_M
  • llama.cppMac, Windows, Linuxllama-server -hf unsloth/Seed-OSS-36B-Instruct-GGUF:UD-IQ1_M
  • Docker Model RunnerMac, Windows, Linuxdocker model run hf.co/unsloth/Seed-OSS-36B-Instruct-GGUF:UD-IQ1_M
On a PC Graphics Card
Video memoryLargest size that fitsDownloadMemory at 8K
6 GBNone worth runningneeds 14 GB
8 GBNone worth runningneeds 14 GB
10 GBNone worth runningneeds 14 GB
11 GBNone worth runningneeds 14 GB
12 GBNone worth runningneeds 14 GB
16 GBUD-IQ3_XXS14 GB16 GB
20 GBUD-Q3_K_XL18 GB19 GB
24 GBUD-Q4_K_XL22 GB23 GB
32 GBQ6_K30 GB31 GB
48 GBUD-Q8_K_XL43 GB44 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-IQ1_MRuns well10 GB5.5
M6 · 24 GBUD-IQ3_XXSRuns well16 GB4.2
M6 · 32 GBIQ4_XSRuns well21 GB3.4
M5 · 16 GBUD-IQ1_MRuns well10 GB5
M5 · 24 GBUD-IQ3_XXSRuns well16 GB3.8
M5 · 32 GBIQ4_XSRuns well21 GB3.1
M5 Pro · 24 GBUD-IQ3_XXSRuns well16 GB7.6
M5 Pro · 48 GBQ4_K_MRuns great23 GB5.7
M5 Pro · 64 GBUD-Q5_K_XLRuns great31 GB5.1
M5 Max · 36 GBUD-Q4_K_XLRuns well23 GB8.4
M5 Max · 48 GBQ4_K_MRuns great23 GB11
M5 Max · 64 GBUD-Q5_K_XLRuns great31 GB9.9
M5 Max · 128 GBQ8_0Runs great58 GB8.7
M5 Ultra · 96 GBQ8_0Runs great43 GB17
M5 Ultra · 256 GBQ8_0Runs great118 GB17
M5 Ultra · 512 GBQ8_0Runs great118 GB17
M4 · 16 GBUD-IQ1_MRuns well10 GB3.8
M4 · 24 GBUD-IQ3_XXSRuns well16 GB3
M4 · 32 GBIQ4_XSRuns well21 GB2.5
M4 Pro · 24 GBUD-IQ3_XXSRuns well16 GB6.5
M4 Pro · 48 GBQ4_K_MRuns great23 GB5
M4 Pro · 64 GBUD-Q5_K_XLRuns great31 GB4.4
M4 Max · 36 GBUD-Q4_K_XLRuns well23 GB7.1
M4 Max · 48 GBQ4_K_MRuns great23 GB9.2
M4 Max · 64 GBUD-Q5_K_XLRuns great31 GB8.3
M4 Max · 96 GBQ8_0Runs great43 GB5.5
M4 Max · 128 GBQ8_0Runs great58 GB7.1
M3 · 8 GBUD-IQ1_SToo big88 GB
M3 · 16 GBUD-IQ1_MRuns well10 GB3.3
M3 · 24 GBUD-IQ3_XXSRuns well16 GB2.5
M3 Pro · 18 GBUD-IQ2_XXSRuns well12 GB4.5
M3 Pro · 36 GBUD-Q4_K_XLRuns well23 GB2.8
M3 Max · 36 GBUD-Q4_K_XLRuns well23 GB5.4
M3 Max · 48 GBQ4_K_MRuns great23 GB7.1
M3 Max · 64 GBUD-Q5_K_XLRuns great31 GB6.4
M3 Max · 96 GBQ8_0Runs great43 GB4.1
M3 Max · 128 GBQ8_0Runs great58 GB5.4
M3 Ultra · 96 GBQ8_0Runs great43 GB10
M3 Ultra · 256 GBQ8_0Runs great118 GB10
M3 Ultra · 512 GBQ8_0Runs great118 GB10
M2 · 8 GBUD-IQ1_SToo big88 GB
M2 · 16 GBUD-IQ1_MRuns well10 GB3.3
M2 · 24 GBUD-IQ3_XXSRuns well16 GB2.5
M2 Pro · 16 GBUD-IQ1_MRuns well10 GB6.2
M2 Pro · 32 GBIQ4_XSRuns well21 GB4
M2 Max · 32 GBIQ4_XSRuns well21 GB7.7
M2 Max · 64 GBUD-Q5_K_XLRuns great31 GB6.4
M2 Max · 96 GBQ8_0Runs great43 GB5.4
M2 Ultra · 64 GBUD-Q5_K_XLRuns great31 GB12
M2 Ultra · 128 GBQ8_0Runs great58 GB10
M2 Ultra · 192 GBQ8_0Runs great118 GB10
M1 · 8 GBUD-IQ1_SToo big88 GB
M1 · 16 GBUD-IQ1_MRuns well10 GB2.2
M1 Pro · 16 GBUD-IQ1_MRuns well10 GB5.9
M1 Pro · 32 GBIQ4_XSRuns well21 GB3.9
M1 Max · 32 GBIQ4_XSRuns well21 GB7.2
M1 Max · 64 GBUD-Q5_K_XLRuns great31 GB6.1
M1 Ultra · 64 GBUD-Q5_K_XLRuns great31 GB11
M1 Ultra · 128 GBQ8_0Runs great58 GB9.5

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

Packaged by

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

Information

Maker
ByteDance
Released
August 2025
Parameters
36B
Longest chat
512K tokens
Reads pictures
No
Licence
apache-2.0
Downloads this month
9,797
Likes
47
Sizes
26, 8.5 to 72 GB
Kind
Maker's release

Badge for the model card

Shows on Seed OSS 36B'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