What hardware do you need to run Qwen3.5 9B?

Qwen3.5 9B at Q4_K_M is 5.7 GB of weights, with a context ceiling of 256k tokens. Not yet rated: released after our last ratings pass. The small end of the current Qwen ladder, which jumps from 9B straight to 27B.

Cheapest machine that runs itMac mini M6, 16GB at $899

Fastest of the ones listedMac Studio M5 Ultra, 96GB — 133 tok/s at 32k context

Honest answer on costPays back in 69 years at 500k tokens a day.

How good is it, really?

On the Artificial Analysis Intelligence Index v4.3 it scores 14 (reasoning mode; 13 without), which puts it in the Below every hosted tier band. Fine for simple, well-specified tasks. Noticeably less capable than anything the big labs sell today. Score source. See the whole table.

Summarising — not rated
Translation — not rated
Everyday coding — not rated
Reasoning & maths — not rated
Agentic work — not rated

What it costs either way

Renting the same model costs $0.08 per million input tokens and $0.13 per million output (OpenRouter, cheapest active endpoint, checked 2026-09-15). Buying a machine only beats that if you use it hard enough, for long enough, that the hardware price divides down below the rental bill.

Machines that run it

Machine	Price	Speed at 32k	Pay-back
Mac mini M6, 16GB	$899	17 tok/s estimated	Pays back in 69 years	Run the numbers
Strix Halo Framework Desktop, 32GB	$1,269	28 tok/s estimated	Pays back in 100 years	Run the numbers
MacBook Air M5 (13-inch), 16GB	$1,299	17 tok/s estimated	Pays back in 99 years	Run the numbers
MacBook Pro M5 (14-inch), 16GB	$1,999	17 tok/s estimated	Pays back in 152 years	Run the numbers
Mac Studio M5 Max, 36GB	$2,499	51 tok/s estimated	Pays back in 183 years	Run the numbers
DGX Spark GB10 Grace Blackwell, 128GB	$4,699	30 tok/s estimated	Pays back in 376 years	Run the numbers

One machine per family, cheapest first. Speeds are measured where a public benchmark exists and estimated from memory bandwidth otherwise; the calculator says which for any configuration.

The specifics

Parameters: 9.7B
Quantisation: Q4_K_M
Weights on disk: 5.7 GB
KV cache: 1.1 GB at 32k context — Hybrid: three linear-attention layers per full-attention layer, so only 8 of 32 hold a growing cache.
Maximum context: 256k tokens (256k)
Licence: Apache 2.0
Sources: source 1