Qwen3.5 122B-A10B Instruct
Qwen3.5 122B-A10B Instruct loads in 72 GB at Q4_K_M and decodes at the speed of its 10B active path, comfortably from 96 GB of memory — Macs included, starting at the MacBook Pro M5 Max 128GB. Below is the memory math we ran for it, the cheapest card that fits, and the fastest machine we track.
How much memory does Qwen3.5 122B-A10B Instruct need?
Qwen3.5 122B-A10B Instruct is a 122B mixture-of-experts model (10B active per token) built for frontier-level reasoning, complex tasks. The weights alone take 72 GB at Q4_K_M; add context and the OS, and it belongs in the 96 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 Qwen3.5 122B-A10B Instruct?
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 Qwen3.5 122B-A10B Instruct need?
72 GB for the Q4_K_M weights, plus ~0.4 GB of KV-cache at 16k context — about 72.4 GB total. Comfortable from 96 GB of VRAM or unified memory.
Does Qwen3.5 122B-A10B Instruct run on a Mac?
Yes — from the MacBook Pro M5 Max 128GB (~19 tok/s est.). Unified memory means the RAM budget is the only limit.
What is the cheapest GPU for Qwen3.5 122B-A10B Instruct?
The AMD Ryzen AI Max+ 395 (Strix Halo) is the cheapest tracked card that runs Qwen3.5 122B-A10B Instruct comfortably — ~9 tok/s est. at ~$3,847 used (as of 2026-07-31).
Cite this page
ModelFit: Qwen3.5 122B-A10B Instruct — specs, memory math and hardware verdicts. https://modelfit.io/models/qwen3.5-122b-a10b/ (dataset updated 2026-09-03, CC BY 4.0).