picksbycard
November 22, 2026No. 204
Home / Will it run? / Qwen3 32B / Radeon RX 7900 XT (used)
Radeon RX 7900 XT (used) · 20GB GDDR6

Can the Radeon RX 7900 XT (used) run Qwen3 32B?

Fits, tight on context at Q4_K_M.

Qwen3 32B is 32.8B parameters. At Q4_K_M the weights are about 18.3 GiB; with the 1.5 GiB runtime reserve that is 19.8 GiB against the card's 20GB, leaving about 0.2 GiB for context.

All three quants

20GB card
QuantWeightsWith reserveVerdictHeadroom
Q4_K_Mthe everyday quant18.3 GiB19.8 GiBFits, tight on context0.2 GiB
Q8_0near-lossless32.4 GiB33.9 GiBOnly with CPU offloadshort 13.9 GiB
FP16full weights61.1 GiB62.6 GiBDoes not fitshort 42.6 GiB

Verdict rules: fits comfortably when weights plus reserve sit within 85% of VRAM; tight when they fit with little left for context; CPU offload when they exceed VRAM by up to 2x (runs, slowly, with layers in system RAM); no beyond that. The KV cache per token depends on the model's architecture and is not modelled here.

The numbers for this pairing

At Q4_K_M the weights take about 18.3 GiB; with the 1.5 GiB we set aside for the runtime that is 19.8 of the card's 20 GB, leaving about 0.2 GB for the context cache.

The cheapest card on our list that runs it comfortably at Q4 is the Radeon RX 7900 XTX at about $1,500, against about $1,018 for this card.

At about $1,018, this card costs roughly $51 per GB of VRAM.

Running it at its full 300 W board power for 4 hours a day would use about 36 kWh a month, about $6.62 at the US average residential rate of 18.4 cents per kWh (2026-05); real use sits below that ceiling.

Prices are the lowest current listings our tracker saw on Nov 22, 2026, since averages include premium variants; launch prices are marked where we have no listing.

Good match. Get the card.

The Radeon RX 7900 XT (used) clears Qwen3 32B at Q4_K_M. Live listings below; this page was rebuilt 2026-11-22.

Amazon links may earn us a commission. Prices move daily; the links open live listings.

About the model

Qwen3 32B by Alibaba Qwen: 32,768-token native context, text, released 2025 under the Apache-2.0. Model card → · Hosted pricing on PicksByModel →

The Silicon Note

One short read on the GPU market, most mornings. No spam, one-click out.