Qwen3 235B A22B
Qwen3 235B A22B loads in 130 GB at Q4_K_M and decodes at the speed of its 22B active path, comfortably from 192 GB of memory — Macs included, starting at the Mac Studio M5 Ultra 256GB. Below is the memory math we ran for it, the cheapest card that fits, and the fastest machine we track.
How much memory does Qwen3 235B A22B need?
Qwen3 235B A22B is a 235B mixture-of-experts model (22B active per token) built for quality, reasoning. The weights alone take 130 GB at Q4_K_M; add context and the OS, and it belongs in the 192 GB memory class. Every number below comes from the ModelFit engine — speeds are estimates, labeled est., because we do not benchmark hardware ourselves.
KV = fp16 estimate (q8_0 cache roughly halves it). "Comfortable" = weights + KV within the engine's tiered budget (~70-85% of memory).
Which machines run Qwen3 235B A22B?
Verdicts come from the same engine that powers our fit checker — memory capacity first, bandwidth for speed. How we compute these numbers · Hugging Face model card
More Qwen models
Frequently asked questions
How much memory does Qwen3 235B A22B need?
130 GB for the Q4_K_M weights, plus ~6.0 GB of KV-cache at 16k context — about 136.0 GB total. Comfortable from 192 GB of VRAM or unified memory.
Does Qwen3 235B A22B run on a Mac?
Yes — from the Mac Studio M5 Ultra 256GB (~14 tok/s est.). Unified memory means the RAM budget is the only limit.
What is the cheapest GPU for Qwen3 235B A22B?
No tracked consumer GPU runs Qwen3 235B A22B comfortably.
Cite this page
ModelFit: Qwen3 235B A22B — specs, memory math and hardware verdicts. https://modelfit.io/models/qwen3-235b-a22b/ (dataset updated 2026-09-03, CC BY 4.0).