Skip to content
Local AI Wizard
Your machine
Theme

Qwen3.5 family: which size to run

2 sizes, 9B–35B parameters · Alibaba

Checked

Sizes

Qwen3.5 sizes
ModelArchitecture
Qwen3.5 9BAlibaba · Apache 2.0 · Mar 2, 20269BDense262,144No LiveBench result for this model in release 2026-06-251595.76
Qwen3.5 35B A3BAlibaba · Apache 2.0 · Feb 24, 202635BMoE · 3B active262,144No LiveBench result for this model in release 2026-06-2535292220.4

Each size is its own model, with its own scores and memory needs. Q4 estimates are planning figures, not measurements; file sizes are published. The Mac or PC column is for a 32 GB or larger machine (8 GB reserve); machines under 32 GB reserve 25% of their memory (2 to 8 GB), so they need less. The publisher lists 0.8B, 2B, 4B, 9B, 27B, 35B-A3B, 122B-A10B, 397B-A17B.

How to read this table

Q4 means compressed to about 4 bits per weight. GGUF (llama.cpp, Ollama, LM Studio) and MLX (Apple silicon) are file formats; the sizes shown are the published 4-bit files. A dash means no value is published or collected for that size; the reason is in the cell's accessible text.

Which size fits my machine

The largest size whose Q4 planning estimate fits what each machine offers. A planning estimate, not a measurement and not a speed.

Largest Qwen3.5 size per machine
MachineLargest Qwen3.5 size with enough memory
Mac mini M2 · 8 GB unified8No size of this family has a Q4 estimate within what this machine offersNo size fits, so there is no estimate to compare
Mac mini M4 · 16 GB unified16Qwen3.5 9B11
Mac mini M4 · 24 GB unified24Qwen3.5 9B13
Mac mini M4 · 32 GB unified32Qwen3.5 9B15
Mac mini M5 Pro · 64 GB unified56Qwen3.5 35B A3B35
Mac Studio M5 Max · 128 GB unified104Qwen3.5 35B A3B35
Mac Studio · M5 Ultra · 36-core CPU / 80-core GPU · 512 GB unified392Qwen3.5 35B A3B35
RTX 4060 · 8 GB VRAM8No size of this family has a Q4 estimate within what this machine offersNo size fits, so there is no estimate to compare
RTX 3060 · 12 GB VRAM12Qwen3.5 9B9
RTX 5080 · 16 GB VRAM16Qwen3.5 9B9
RTX 4090 · 24 GB VRAM24Qwen3.5 9B9
RTX 5090 · 32 GB VRAM32Qwen3.5 35B A3B29
DGX H100 · 8 × 80 GB640Qwen3.5 35B A3B43

A dash means no size of this family fits that machine. The requirement is the size's Q4 estimate with that machine's own reserve: a graphics card reserves 2 GB, as in the sizes table; a Mac or PC reserves 25% of its memory (2 to 8 GB), so below 32 GB its figure is lower than the sizes table's. A server profile reserves 2 GB per graphics card, so its figure is higher.

Smallest machine for each size

The profile with the least memory that still has enough for the size's Q4 estimate, from the hardware catalog.

Smallest machine per Qwen3.5 size
SizeSmallest Mac or PC-memory profileSmallest graphics-card profile
Qwen3.5 9BMac mini 2020 · M1 · 16 GBRTX 3060 · 12 GB
Qwen3.5 35B A3BMac Studio 2026 · M5 Max · 36 GBRTX 5090 · 32 GB

A dash means no profile of that kind in the catalog has enough memory for the size.

Memory by file precision

The same planning formula at three precisions (0.625, 1.125 and 2 bytes per parameter, plus working room and the system reserve). Estimates, not measurements.

Qwen3.5 memory estimate by precision
Size
Qwen3.5 9B15921153024
Qwen3.5 35B A3B352956509286

Family tree

Family, sizes and, for each size, the two 4-bit files the tables use (GGUF Q4_K_M and MLX 4-bit) when one was collected. Other precisions are not drawn here.

Qwen3.5 family: sizes and published filesSwipe sideways to see the files, or read the list below.
Qwen3.5 family: sizes and published filesFamilySizesFilesQwen3.5Alibaba9BGGUF Q4_K_M5.7 GBMLX 4-bit6 GB35B A3B3B activeGGUF Q4_K_M22 GBMLX 4-bit20.4 GB

The same tree as a list:

  • Qwen3.5 · Alibaba
    • 9B
      • GGUF Q4_K_M · 5.7 GB · unsloth/Qwen3.5-9B-GGUF, read Oct 23, 2026
      • MLX 4-bit · 6 GB · lmstudio-community/Qwen3.5-9B-MLX-4bit, read Oct 23, 2026
    • 35B A3B · 3B active
      • GGUF Q4_K_M · 22 GB · unsloth/Qwen3.5-35B-A3B-GGUF, read Oct 23, 2026
      • MLX 4-bit · 20.4 GB · mlx-community/Qwen3.5-35B-A3B-4bit, read Oct 23, 2026

Learn more

Sources

Where the family and its sizes are named, and where each number comes from.

  • QwenLM/Qwen3.5 README · “2026-02-16: We release Qwen3.5. The first release includes a 397B-A17B MoE model.” · read Oct 17, 2026
  • QwenLM/Qwen3.5 README · “2026-03-02: Qwen3.5-9B, Qwen3.5-4B, Qwen3.5-2B, and Qwen3.5-0.8B are now available on Hugging Face Hub” · read Oct 17, 2026
  • QwenLM/Qwen3.5 README · “2026-02-24: Qwen3.5-122B-A10B, Qwen3.5-35B-A3B, and Qwen3.5-27B are released.” · read Oct 17, 2026
  • Qwen3.5 9B model card
  • Qwen3.5 35B A3B model card
  • LiveBench 2026-06-25 · the only quality source on this page; a dash means the release does not include the model · snapshot collected Sep 26, 2026
  • Published file sizes: Hugging Face repositories, read Oct 23, 2026
  • Methodology · how the Q4 planning estimate is computed · Learn