▲ local intelligence

can it run

Can the NVIDIA GeForce RTX 4080 run SDXL SDXL base?

FULL

8.9 GB needed of 16 GB usable — headroom for context

16GB · Ada · model file 6.9GB · SDXL base

~48–79 tok/s

computed roofline: 716.8 GB/s × 0.46–0.76 efficiency window / 6.9GB (Q4_K_M) — bands, never points.

Run SDXL SDXL base on instead

→ Tesla M60 (16GB, used) — cheapest catalog machine that runs SDXL SDXL base fully ($35)

→ rtx-4090 — the next step up from NVIDIA GeForce RTX 4080

→ every model the NVIDIA GeForce RTX 4080 can run — full list

→explore the full catalog91 machines indexed · live prices · what each one can run
2026-10-01 · ← all models on the NVIDIA GeForce RTX 4080 · fit = weights + 4k KV vs usable memory