Cheapest Mac to run Qwen3.5 35B-A3B
35B parameters (3B active) · quality index 35 · coding 40 · every Apple Silicon Mac ever sold, compared. It is a mixture-of-experts model: all 35B parameters must sit in memory, but only 3B compute per token, which is why it is faster than dense models of similar size.
The cheapest Mac that runs Qwen3.5 35B-A3B comfortably is a used Mac mini M4 10c/10c 32GB (2024) at about $769 EST. on the used market — running Q4_K_M quantization at roughly 33 tok/s with up to 60K tokens of context.
Every Mac that runs it, by used price
| Machine | Used price | Runs at | Est. speed | Max context |
|---|---|---|---|---|
| Mac mini M4 10c/10c 32GB (2024) | $769 EST. | Q4_K_M | 33 tok/s | 60K tokens |
| Mac mini M2 Pro 10c/16c 32GB (2023) | $1,062 EST. | Q4_K_M | 55 tok/s | 60K tokens |
| Mac Studio M1 Max 10c/24c 32GB (2022) | $1,229 EST. | Q4_K_M | 110 tok/s | 60K tokens |
| Mac Studio M2 Max 12c/30c 32GB (2023) | $1,249 EST. | Q4_K_M | 110 tok/s | 60K tokens |
| Mac mini M2 Pro 12c/19c 32GB (2023) | $1,249 EST. | Q4_K_M | 55 tok/s | 60K tokens |
| Mac Studio M1 Max 10c/32c 32GB (2022) | $1,352 EST. | Q4_K_M | 110 tok/s | 60K tokens |
| Mac Studio M2 Max 12c/38c 32GB (2023) | $1,374 EST. | Q4_K_M | 110 tok/s | 60K tokens |
| Mac Studio M1 Max 10c/24c 64GB (2022) | $1,475 EST. | Q8_0 | 60 tok/s | 177K tokens |
| MacBook Pro M1 Pro 10c/16c 32GB (2021) | $1,484 EST. | Q4_K_M | 55 tok/s | 60K tokens |
| Mac Studio M2 Max 12c/30c 64GB (2023) | $1,499 EST. | Q8_0 | 60 tok/s | 177K tokens |
See all 55 machines, other price bases, and live currency conversion →
Run it
ollama run qwen3.5:35b pulls the default (≈Q4) build once you have Ollama installed.
Or skip the hardware
Cloud APIs serve Qwen3.5 35B-A3B at about $0.15 per million output tokens. The main site's break-even solver computes the daily usage where owning a Mac becomes cheaper than renting.
Open the interactive guide — speed simulator, TCO, all 75 machines →