Outmake Models
Models

Qwen3.5 9B

Alibaba Qwen · March 2026

GetHugging Face

A chat model Outmake recommends for a 16 GB Mac that reads images as well as text.

Rating, estimated
3.4
Downloads
2.6M
Parameters
9B
Context
256K
Download
4.1 to 19 GB
Commercial use
Yes

Runs great on M4 · 16 GB

Outmake picks the Q4_K_M size: a 6.6 GB download that uses 7.2 GB of memory, and writes about 13 words (17 tokens) a second (measured).

About

Vision and language were trained together on image and text input combined.

How it compares

Outmake estimates Qwen3.5 4B writes about 1.5 times as fast on a 16 GB Mac, and it reads pictures too.

By Outmake's estimate for a 16 GB Mac, Gemma 4 12B and Gemma 4 E4B write at about the same speed.

On Hugging Face last month, it was the 3rd most downloaded of the 7 chat models Outmake recommends for a 16 GB Mac.

Good to know

The maker says its reinforcement learning ran across million-agent environments with increasingly complex tasks.

Sources: Qwen/Qwen3.5-9B model card, Outmake's speed estimates, LM Arena and Hugging Face. Every line comes from one of these.

Good at

  • Reading images as well as text
  • Text in 201 languages and dialects
  • Calling tools in Outmake
  • Thinking before answering

Licence

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

Qwen3.5 9B runs well from a Mac with an M1 and 8 GB of memory, at UD-Q2_K_XL, writing about 6.9 words (9.3 tokens) a second. 59 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
Q3_K_MSmaller5.6 GB5.9 GB6.3 GBOpen
Q4_K_MBalanced6.6 GB6.7 GB7.2 GBOpen
Q6_KBest8.4 GB8.3 GB8.7 GBOpen
All 22 sizes
SizeDownloadMemory at 8KMemory at 32KOpen
UD-IQ2_XXS4.1 GB4.4 GBestimated4.8 GBOpen
UD-IQ2_M4.6 GB4.8 GBestimated5.3 GBOpen
UD-IQ3_XXS4.9 GB5.2 GBestimated5.6 GBOpen
UD-Q2_K_XL5.0 GB5.3 GBestimated5.7 GBOpen
Q3_K_S5.2 GB5.5 GBestimated5.9 GBOpen
UD-Q3_K_XL6.0 GB6.2 GBestimated6.7 GBOpen
IQ4_XS6.1 GB6.4 GBestimated6.8 GBOpen
IQ4_NL6.3 GB6.4 GBestimated6.9 GBOpen
Q4_06.3 GB6.4 GBestimated6.9 GBOpen
Q4_K_S6.3 GB6.4 GBestimated6.9 GBOpen
Q4_16.8 GB6.9 GBestimated7.3 GBOpen
UD-Q4_K_XL6.9 GB7.0 GBestimated7.4 GBOpen
Q5_K_S7.3 GB7.4 GBestimated7.8 GBOpen
Q5_K_M7.5 GB7.5 GBestimated7.8 GBOpen
UD-Q5_K_XL7.7 GB7.7 GBestimated8.0 GBOpen
UD-Q6_K_XL9.7 GB9.7 GBestimated10.0 GBOpen
Q8_010 GB10 GBestimated11 GBOpen
UD-Q8_K_XL14 GB14 GBestimated14 GBOpen
BF1619 GB19 GBestimated19 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

Qwen3.5 9B is an open model, so any app that runs GGUF files can run it. These take the Q4_K_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/Qwen3.5-9B-GGUF:Q4_K_M
  • llama.cppMac, Windows, Linuxllama-server -hf unsloth/Qwen3.5-9B-GGUF:Q4_K_M
  • Docker Model RunnerMac, Windows, Linuxdocker model run hf.co/unsloth/Qwen3.5-9B-GGUF:Q4_K_M
On a PC Graphics Card
Video memoryLargest size that fitsDownloadMemory at 8K
6 GBQ3_K_M5.6 GB5.9 GB
8 GBUD-Q5_K_XL7.7 GB7.7 GB
10 GBUD-Q6_K_XL9.7 GB9.7 GB
11 GBQ8_010 GB10 GB
12 GBQ8_010 GB10 GB
16 GBUD-Q8_K_XL14 GB14 GB
20 GBBF1619 GB19 GB
24 GBBF1619 GB19 GB
32 GBBF1619 GB19 GB
48 GBBF1619 GB19 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
MacSizeFitMemoryWritesReads
M6 · 16 GBQ4_K_MRuns great7.2 GB14699
M6 · 24 GBQ6_KRuns great11 GB11699
M6 · 32 GBQ6_KRuns great14 GB11699
M5 · 16 GBQ4_K_MRuns great7.2 GB12258
M5 · 24 GBQ6_KRuns great11 GB10258
M5 · 32 GBQ6_KRuns great14 GB10258
M5 Pro · 24 GBQ6_KRuns great11 GB20516
M5 Pro · 48 GBQ6_KRuns great14 GB20516
M5 Pro · 64 GBQ6_KRuns great14 GB20645
M5 Max · 36 GBQ6_KRuns great14 GB301032
M5 Max · 48 GBQ6_KRuns great14 GB391290
M5 Max · 64 GBQ6_KRuns great14 GB391290
M5 Max · 128 GBQ6_KRuns great14 GB391290
M5 Ultra · 96 GBQ6_KRuns great14 GB732064
M5 Ultra · 256 GBQ6_KRuns great14 GB732580
M5 Ultra · 512 GBQ6_KRuns great14 GB732580
M4 · 16 GBQ4_K_MRuns great7.2 GB13measured, 8 answers in use, drafting on, early in a chat150
M4 · 24 GBQ6_KRuns great11 GB7.787
M4 · 32 GBQ6_KRuns great14 GB7.787
M4 Pro · 24 GBQ6_KRuns great11 GB16173
M4 Pro · 48 GBQ6_KRuns great14 GB16173
M4 Pro · 64 GBQ6_KRuns great14 GB16173
M4 Max · 36 GBQ6_KRuns great14 GB23347
M4 Max · 48 GBQ6_KRuns great14 GB29433
M4 Max · 64 GBQ6_KRuns great14 GB29433
M4 Max · 96 GBQ6_KRuns great14 GB23347
M4 Max · 128 GBQ6_KRuns great14 GB29433
M3 · 8 GBUD-Q2_K_XLRuns well5.3 GB1072
M3 · 16 GBQ4_K_MRuns great7.2 GB8.172
M3 · 24 GBQ6_KRuns great11 GB6.672
M3 Pro · 18 GBQ6_KRuns great8.7 GB9.6126
M3 Pro · 36 GBQ6_KRuns great14 GB9.6126
M3 Max · 36 GBQ6_KRuns great14 GB18269
M3 Max · 48 GBQ6_KRuns great14 GB23359
M3 Max · 64 GBQ6_KRuns great14 GB23359
M3 Max · 96 GBQ6_KRuns great14 GB18269
M3 Max · 128 GBQ6_KRuns great14 GB23359
M3 Ultra · 96 GBQ6_KRuns great14 GB41538
M3 Ultra · 256 GBQ6_KRuns great14 GB41538
M3 Ultra · 512 GBQ6_KRuns great14 GB41718
M2 · 8 GBUD-Q2_K_XLRuns well5.3 GB1070
M2 · 16 GBQ4_K_MRuns great7.2 GB8.170
M2 · 24 GBQ6_KRuns great11 GB6.570
M2 Pro · 16 GBQ4_K_MRuns great7.2 GB15141
M2 Pro · 32 GBQ6_KRuns great14 GB13141
M2 Max · 32 GBQ6_KRuns great14 GB23264
M2 Max · 64 GBQ6_KRuns great14 GB23264
M2 Max · 96 GBQ6_KRuns great14 GB23335
M2 Ultra · 64 GBQ6_KRuns great14 GB40528
M2 Ultra · 128 GBQ6_KRuns great14 GB40528
M2 Ultra · 192 GBQ6_KRuns great14 GB40528
M1 · 8 GBUD-Q2_K_XLRuns well5.3 GB6.957
M1 · 16 GBQ4_K_MRuns great7.2 GB5.557
M1 Pro · 16 GBQ4_K_MRuns great7.2 GB14114
M1 Pro · 32 GBQ6_KRuns great14 GB12114
M1 Max · 32 GBQ6_KRuns great14 GB21196
M1 Max · 64 GBQ6_KRuns great14 GB21196
M1 Ultra · 64 GBQ6_KRuns great14 GB34392
M1 Ultra · 128 GBQ6_KRuns great14 GB34392

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 Qwen3.5 9B'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/Qwen3.5-9B-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