How much memory does gpt-oss-20b at Q8_0 need?

24GB. That is 22GB of weights at 8.5 bits per weight, 0.4GB of K/V cache at 8K context, and 1.2GB of runtime overhead.

Which machine is cheapest for gpt-oss-20b at Q8_0?

Mac mini, M6, 32GB, $1,499 at Apple. It fits with 5GB of headroom and decodes at an est. 22.2 tok/s.

Cheapest box to run gpt-oss-20b at Q8_0

OpenAI’s gpt-oss-20b is 21B parameters, 3.6B of them read per token. The number below is computed from that and each vendor’s published specs, not asserted. Switch the quant or the context and it recomputes in place.

Fit finderprices checked 10 hours ago

Quant

Context

gpt-oss-20b at Q8_0 needs 24GB.

Cheapest Config that fits: Mac mini, M6, 32GB. Fits, 5GB headroom. Est. 22.2 tok/s.

Best offer today: $1,499 at Apple — in stock, checked 10 hours ago.

Collecting price history — 2 of 90 days. No timing verdict yet.

Offers — Mac mini, M6, 32GB, 512GB

Applelist pricein stockchecked 10 hours ago$1,499
Expercomout of stockchecked 10 hours ago$1,499
B&H Photoout of stockchecked 10 hours ago$1,599

Mac mini, M6, 16GB$1,099
Won't fit−11GB headroom
Mac mini, M6, 24GB$1,299
Won't fit−3GB headroom
Mac mini, M6, 32GB$1,499
Fits5GB headroomest. 22.2 tok/s
Mac mini, M5 Pro, 24GB$1,699
Won't fit−3GB headroom
Mac mini, M5 Pro, 48GB$2,499
Fits20GB headroomest. 40.1 tok/s
Mac Studio, M5 Max, 36GB$2,499
Fits9GB headroomest. 80.3 tok/s
Mac mini, M5 Pro, 64GB$2,899
Fits35GB headroomest. 40.1 tok/s
GeForce RTX 5090, 32GB$2,999.99+host
Fits7GB headroomest. 234 tok/s
Mac Studio, M5 Max, 48GB$3,099
Fits20GB headroomest. 80.3 tok/s
Mac Studio, M5 Max, 64GB$3,499
Fits35GB headroomest. 80.3 tok/s
Ascent GX10, GB10 Grace Blackwell, 128GB$3,999.99
Fits96GB headroomest. 35.7 tok/s
DGX Spark, GB10 Grace Blackwell, 128GB$4,699
Fits96GB headroomest. 35.7 tok/s
Mac Studio, M5 Max, 128GB$5,099
Fits96GB headroomest. 80.3 tok/s
Mac Studio, M5 Ultra, 96GB$5,499
Fits64GB headroomest. 157 tok/s
Mac Studio, M5 Ultra, 256GB$9,499
Fits224GB headroomest. 157 tok/s
Mac Studio, M5 Ultra, 512GB—
Fits480GB headroomest. 157 tok/s

Config	Memory	Price	$/GB	Fit	Headroom	Est. speed
Mac mini, M6	16GB	$1,099	$69	Won't fit	−11GB	—
Mac mini, M6	24GB	$1,299	$54	Won't fit	−3GB	—
Mac mini, M6	32GB	$1,499	$47	Fits	5GB	22.2 tok/s
Mac mini, M5 Pro	24GB	$1,699	$71	Won't fit	−3GB	—
Mac mini, M5 Pro	48GB	$2,499	$52	Fits	20GB	40.1 tok/s
Mac Studio, M5 Max	36GB	$2,499	$69	Fits	9GB	80.3 tok/s
Mac mini, M5 Pro	64GB	$2,899	$45	Fits	35GB	40.1 tok/s
GeForce RTX 5090	32GB	$2,999.99+host	$94	Fits	7GB	234 tok/s
Mac Studio, M5 Max	48GB	$3,099	$65	Fits	20GB	80.3 tok/s
Mac Studio, M5 Max	64GB	$3,499	$55	Fits	35GB	80.3 tok/s
Ascent GX10, GB10 Grace Blackwell	128GB	$3,999.99	$31	Fits	96GB	35.7 tok/s
DGX Spark, GB10 Grace Blackwell	128GB	$4,699	$37	Fits	96GB	35.7 tok/s
Mac Studio, M5 Max	128GB	$5,099	$40	Fits	96GB	80.3 tok/s
Mac Studio, M5 Ultra	96GB	$5,499	$57	Fits	64GB	157 tok/s
Mac Studio, M5 Ultra	256GB	$9,499	$37	Fits	224GB	157 tok/s
Mac Studio, M5 Ultra	512GB	—	—	Fits	480GB	157 tok/s

GeForce RTX 5090 is a card, not a computer: every price shown for it is the cheapest listing we found plus $1,000 for the minimum host machine.

memory = 21B × 1.063 B/param (8.5 bits) + 0.049 GB/1K × 8K K/V + 1.2 GB runtime
speed = bandwidth × 0.5 ÷ (3.6B active × 1.063 B/param) = 3.8 GB read per token
usable = memory − operating system reserve (8% on a Mac, 6% on a DGX Spark, 4% on a card)

Bandwidth is the vendor’s published figure. Every speed is an estimate and says so. Methodology.

What else Mac mini M6 32GB runs

Qwen3-32BQ5_K_M, est. 4.9 tok/s
Gemma-3-27BQ6_K, est. 5.0 tok/s
gpt-oss-20bQ8_0, est. 22.2 tok/s

Best quant that still fits at 8K context, largest model first. Full list on the Config page.

Bandwidth 170GB/s, usable memory 29GB of 32GB.