local intelligence

can it run

Can the RTX 3090 (24GB, used) run FLUX dev fp8?

STRETCH

Only with compromises. 24GB is below the 29GB working set — expect a lower quant, offloading, or heavy swap. Possible, not pleasant.

~19–31 tok/s

weights 23GB + KV cache 2GB at 4k context vs ~24GB usable → tight. 25.0 of 24 GB usable — barely over; real with q8 KV cache (halves KV) or shorter context

computed roofline: 936 GB/s × 0.46–0.76 efficiency window / 23.0GB (Q4_K_M) — bands, never points.

explore the full catalog84 machines indexed · live prices · what each one can run

Specs

model file23GB (GGUF Q4_K_M)
minimum memory29GB working set
this machine24GB GDDR6X, 936 GB/s
paramsdev fp8
licenseflux-dev-nc / apache schnell
price$1100

Run it

model files: hugging face ↗

easy run: no stock runtime yet

Alternatives

Mac mini M5 Pro (64GB) — cheapest machine that runs it ($1669)

ASUS ExpertCenter Pro ET900N G3 (GB300) — this model on the biggest machine

rtx-5090 — the next step up from this machine's GPU

Buy

NVIDIA (used market) ↗ · Amazon ↗

2026-09-17 · ← full hardware catalog · prices verified at source, may drift