Outmake Models
Models

Qwen3.6 35B

Alibaba Qwen · April 2026

GetHugging Face

Coding help that can read images, and the fastest chat model Outmake recommends for a 24 GB Mac.

Rating
3.6
Downloads
3.9M
Parameters
36B, 3B active
Context
256K
Download
11 to 70 GB
Commercial use
Yes

Tight fit on M4 · 16 GB

Outmake picks the UD-IQ2_XXS size: a 12 GB download that uses 12 GB of memory, and writes about 23 words (31 tokens) a second.

About

Qwen built it from community feedback, with a focus on stability and real-world coding.

How it compares

Outmake estimates it writes about 1.4 times as fast as Gemma 4 26B on a 24 GB Mac.

Of the 8 chat models Outmake recommends for a 24 GB Mac, it is one of 6 that read pictures.

Of those 8 chat models, it was the 3rd most downloaded on Hugging Face last month, behind Qwen3.8 27B and Gemma 4 12B.

Good to know

It is the first open model in Qwen's 3.6 line, which follows the Qwen3.5 series from February.

Only about a twelfth of the model is active at any one time.

Sources: Qwen/Qwen3.6-35B-A3B model card, Outmake's speed estimates, LM Arena and Hugging Face. Every line comes from one of these.

Good at

  • Questions across a code project
  • Frontend coding work

Licence

Apache 2.0. Commercial use: yes. The maker's page has the full terms.

Qwen3.6 35B runs well from a Mac with an M3 Pro and 18 GB of memory, at UD-IQ2_XXS, writing about 28 words (38 tokens) a second. 48 of 59 Mac configurations run it well.

Compare with

The nearest models in size, judged for the same Mac.

Sizes

SizeDownloadMemory at 8KMemory at 32KOpen
UD-IQ4_XSSmaller19 GB19 GB19 GBOpen
UD-Q4_K_XLBalanced23 GB23 GB24 GBOpen
UD-Q6_KBest30 GB30 GB31 GBOpen
All 23 sizes
SizeDownloadMemory at 8KMemory at 32KOpen
UD-IQ1_M11 GB11 GBestimated11 GBOpen
UD-IQ2_XXS12 GB12 GBestimated12 GBOpen
UD-IQ2_M12 GB13 GBestimated13 GBOpen
UD-Q2_K_XL13 GB13 GBestimated14 GBOpen
UD-IQ3_XXS14 GB14 GBestimated14 GBOpen
UD-IQ3_S15 GB15 GBestimated15 GBOpen
UD-Q3_K_S16 GB16 GBestimated17 GBOpen
UD-Q3_K_M17 GB18 GBestimated18 GBOpen
UD-Q3_K_XL18 GB18 GBestimated18 GBOpen
UD-IQ4_NL19 GB19 GBestimated19 GBOpen
UD-Q4_K_S22 GB22 GBestimated22 GBOpen
MXFP4_MOE23 GB23 GBestimated23 GBOpen
UD-Q4_K_M23 GB23 GBestimated23 GBOpen
UD-Q5_K_S26 GB26 GBestimated26 GBOpen
UD-Q5_K_M27 GB27 GBestimated28 GBOpen
UD-Q5_K_XL27 GB28 GBestimated28 GBOpen
UD-Q6_K_XL33 GB33 GBestimated33 GBOpen
Q8_038 GB38 GBestimated38 GBOpen
UD-Q8_K_XL39 GB39 GBestimated40 GBOpen
BF1670 GB70 GBestimated71 GBOpen

Memory is what llama.cpp's own calculator reported for each size when the catalog was built, with the app's settings. Smaller, Balanced and Best are the three sizes the store names.

Run It Anywhere

Qwen3.6 35B is an open model, so any app that runs GGUF files can run it. These take the UD-Q4_K_XL size from unsloth straight from Hugging Face.

  • LM StudioMac, Windows, LinuxOpen
  • JanMac, Windows, LinuxOpen
  • MstyMac, Windows, LinuxOpen
  • OllamaMac, Windows, Linuxollama run hf.co/unsloth/Qwen3.6-35B-A3B-GGUF:UD-Q4_K_XL
  • llama.cppMac, Windows, Linuxllama-server -hf unsloth/Qwen3.6-35B-A3B-GGUF:UD-Q4_K_XL
  • Docker Model RunnerMac, Windows, Linuxdocker model run hf.co/unsloth/Qwen3.6-35B-A3B-GGUF:UD-Q4_K_XL
On a PC Graphics Card
Video memoryLargest size that fitsDownloadMemory at 8K
6 GBNone worth runningneeds 13 GB
8 GBNone worth runningneeds 13 GB
10 GBNone worth runningneeds 13 GB
11 GBNone worth runningneeds 13 GB
12 GBNone worth runningneeds 13 GB
16 GBUD-IQ3_S15 GB15 GB
20 GBUD-IQ4_NL19 GB19 GB
24 GBUD-Q4_K_XL23 GB23 GB
32 GBUD-Q6_K30 GB30 GB
48 GBUD-Q8_K_XL39 GB39 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.

Tests

KnowledgeMMLU-Pro
85.2%
ScienceGPQA Diamond
86%
Expert questionsHumanity's Last Exam
21.4%
Software fixesSWE-bench Verified
73.4%
Terminal tasksTerminal-Bench 2.0
51.5%
MathsAIME 2026
92.7%
VisionMMMU-Pro
75.3%

MMLU-Pro: graduate-level questions in 14 subjects. GPQA Diamond: questions PhD scientists write for each other. Humanity's Last Exam: questions written to be beyond today's models. SWE-bench Verified: real GitHub issues fixed so the project's tests pass. Terminal-Bench 2.0: jobs finished alone in a terminal. AIME 2026: this year's AIME competition problems. MMMU-Pro: college-level questions about images. 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.

On every Mac
MacSizeFitMemoryWritesReads
M6 · 16 GBUD-IQ2_XXSTight fit12 GB371467
M6 · 24 GBUD-IQ3_SRuns well15 GB331467
M6 · 32 GBUD-IQ4_XSRuns well20 GB281467
M5 · 16 GBUD-IQ2_XXSTight fit12 GB36542
M5 · 24 GBUD-IQ3_SRuns well15 GB32542
M5 · 32 GBUD-IQ4_XSRuns well20 GB27542
M5 Pro · 24 GBUD-IQ3_SRuns well15 GB571083
M5 Pro · 48 GBUD-IQ4_XSRuns great20 GB491083
M5 Pro · 64 GBUD-Q6_KRuns great31 GB351354
M5 Max · 36 GBUD-Q4_K_XLRuns well24 GB592167
M5 Max · 48 GBUD-IQ4_XSRuns great20 GB842709
M5 Max · 64 GBUD-Q6_KRuns great31 GB632709
M5 Max · 128 GBUD-Q6_KRuns great32 GB632709
M5 Ultra · 96 GBUD-Q6_KRuns great32 GB1044334
M5 Ultra · 256 GBUD-Q6_KRuns great32 GB1045417
M5 Ultra · 512 GBUD-Q6_KRuns great32 GB1045417
M4 · 16 GBUD-IQ2_XXSTight fit12 GB23182
M4 · 24 GBUD-IQ3_SRuns well15 GB21182
M4 · 32 GBUD-IQ4_XSRuns well20 GB18182
M4 Pro · 24 GBUD-IQ3_SRuns well15 GB35364
M4 Pro · 48 GBUD-IQ4_XSRuns great20 GB31364
M4 Pro · 64 GBUD-Q6_KRuns great31 GB25364
M4 Max · 36 GBUD-Q4_K_XLRuns well24 GB36728
M4 Max · 48 GBUD-IQ4_XSRuns great20 GB44910
M4 Max · 64 GBUD-Q6_KRuns great31 GB37910
M4 Max · 96 GBUD-Q6_KRuns great32 GB32728
M4 Max · 128 GBUD-Q6_KRuns great32 GB37910
M3 · 8 GBUD-IQ4_XSToo big20 GB
M3 · 16 GBUD-IQ2_XXSTight fit12 GB21151
M3 · 24 GBUD-IQ3_SRuns well15 GB19151
M3 Pro · 18 GBUD-IQ2_XXSRuns well12 GB28264
M3 Pro · 36 GBUD-Q4_K_XLRuns well24 GB20264
M3 Max · 36 GBUD-Q4_K_XLRuns well24 GB32565
M3 Max · 48 GBUD-IQ4_XSRuns great20 GB42753
M3 Max · 64 GBUD-Q6_KRuns great31 GB34753
M3 Max · 96 GBUD-Q6_KRuns great32 GB28565
M3 Max · 128 GBUD-Q6_KRuns great32 GB34753
M3 Ultra · 96 GBUD-Q6_KRuns great32 GB491130
M3 Ultra · 256 GBUD-Q6_KRuns great32 GB491130
M3 Ultra · 512 GBUD-Q6_KRuns great32 GB491507
M2 · 8 GBUD-IQ4_XSToo big20 GB
M2 · 16 GBUD-IQ2_XXSTight fit12 GB21148
M2 · 24 GBUD-IQ3_SRuns well15 GB19148
M2 Pro · 16 GBUD-IQ2_XXSTight fit12 GB34296
M2 Pro · 32 GBUD-IQ4_XSRuns well20 GB27296
M2 Max · 32 GBUD-IQ4_XSRuns well20 GB41555
M2 Max · 64 GBUD-Q6_KRuns great31 GB33555
M2 Max · 96 GBUD-Q6_KRuns great32 GB33703
M2 Ultra · 64 GBUD-Q6_KRuns great31 GB481109
M2 Ultra · 128 GBUD-Q6_KRuns great32 GB481109
M2 Ultra · 192 GBUD-Q6_KRuns great32 GB481109
M1 · 8 GBUD-IQ4_XSToo big20 GB
M1 · 16 GBUD-IQ2_XXSTight fit12 GB14120
M1 Pro · 16 GBUD-IQ2_XXSTight fit12 GB27240
M1 Pro · 32 GBUD-IQ4_XSRuns well20 GB23240
M1 Max · 32 GBUD-IQ4_XSRuns well20 GB31411
M1 Max · 64 GBUD-Q6_KRuns great31 GB27411
M1 Ultra · 64 GBUD-Q6_KRuns great31 GB35822
M1 Ultra · 128 GBUD-Q6_KRuns great32 GB35822

Writes and reads are words a second (a token is about three quarters of a word), 4K tokens into a chat, estimated unless marked measured. The size is the one the store offers that Mac. Constants fitted October 11, 2026; how, and how wrong.

Badge for the model card

Shows on Qwen3.6 35B's Hugging Face card which Macs run it well, linked to this page. It updates every morning.

Runs on Mac badge

Sizes and files are read from unsloth/Qwen3.6-35B-A3B-GGUF on Hugging Face and checked every morning. LM Arena has not rated this model, so the rating is Outmake's estimate from its size and release date; how the estimate is made. Catalog of November 4, 2026.

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