Skip to content
local-ai

Qwen3.6-35B-A3B

Alibaba · Vision-language · MoE 35B (A3B) · 262k context · Released 16 April 2026

Commercial use permitted Text + Image + Video Text Open weights Apple Silicon

Qwen3.6-35B-A3B is a sparse mixture-of-experts with 256 experts, eight routed plus one shared active per token, taking text, images and video and producing text, with a 262k context extensible towards a million tokens. Its sparsity makes it fast for its size, and it is a favourite quick daily driver on Apple Silicon and other unified-memory machines. It fits a 24GB card at a 4-bit quantisation, with the usual mixture-of-experts caveat that the memory footprint is set by the 35 billion total parameters, not the 3 billion active.

Strengths

  • Apache 2.0, freely usable commercially with no territory restriction
  • Sparse and fast: only about 3 billion of 35 billion parameters compute per token
  • Fits a 24GB card at a 4-bit quantisation, quick on unified memory
  • Multimodal input with a 262k context

Weaknesses

  • Memory is set by the 35B total, not the 3B active, a common point of confusion
  • Superseded for raw quality by the newer Qwen3.8 line
  • Image support in local runtimes is less mature than text-only use
  • A mid-sized model still trails the largest open models on hard tasks

Hardware requirements

Quantisation Approx. VRAM Notes
Q4_K_M ~20GB Fits a 24GB card. Driven by the 35B total, not the active count. Approximate
Q8_0 ~37GB Near-lossless, needs a 48GB card or more. Approximate
BF16 ~70GB Full precision, multi-GPU or server territory. Approximate

Optimised builds available for Apple Silicon.

What you'd need to run this

Roughly what a machine to run this would need, at up to three levels of quality. Memory is the deciding factor.

Minimum to run it

Q4_K_M · ~20GB needed

One 24GB GPU

NVIDIA Tesla P40

or a Mac or mini-PC with unified memory, if you prefer no discrete GPU, Mac mini M5 Pro .

32–64GB of system RAM alongside the card.

around £700–£1,100

What else 24GB runs →

For good quality

Q8_0 · ~37GB needed

64GB of unified memory

Mac mini M4 Pro

or a single high-memory workstation card, NVIDIA RTX 6000 Ada Generation .

Unified memory is shared with the model, so it is already counted above.

around £1,100–£2,000

See the Affordable Mac build build What else 64GB runs →

Best quality

BF16 · ~70GB needed

96GB of unified memory

AMD Ryzen AI Max+ 395 (Strix Halo)

or an 80GB-class data-centre card, which is usually rented by the hour, NVIDIA RTX PRO 6000 Blackwell .

Unified memory is shared with the model, so it is already counted above.

around £1,100–£2,000

See the Quiet large-model box build What else 96GB runs →

Licence

Apache 2.0 — read the licence

Qwen3.6-35B-A3B: common questions

What hardware do I need to run Qwen3.6-35B-A3B?
At its most compressed (Q4_K_M) it needs roughly 20GB of VRAM, and about 24GB for good quality. VRAM figures are approximate and depend on context length and settings.
Is Qwen3.6-35B-A3B free for commercial use?
Yes. Qwen3.6-35B-A3B is licensed under Apache 2.0, which permits commercial use with no meaningful conditions.
Can I run Qwen3.6-35B-A3B on Apple Silicon?
Yes. Qwen3.6-35B-A3B has builds optimised for Apple Silicon, through MLX or GGUF on a Mac.
What is Qwen3.6-35B-A3B's context window?
Qwen3.6-35B-A3B has a context window of 262,144 tokens, about 262k.

Availability

Recommended for

  • A fast general and multimodal model on a 24GB card
  • Quick daily use on Apple Silicon or other unified memory
  • Long-context work where speed matters

Related models

Run it with

Glossary

Our coverage

Catalogue entry last verified 2 October 2026. Specifications change; verify anything you are about to spend money on.