local intelligence

can it run

Can the NVIDIA H100 NVL 94GB run devstral 24B?

FULL

15.2 GB needed of 94 GB usable — headroom for context

94GB · Hopper · model file 14.6GB · 24B

~120–200 tok/s

computed roofline: 3900 GB/s × 0.46–0.76 efficiency window / 14.6GB (Q4_K_M) — bands, never points.

Run devstral 24B on instead

Tesla M40 (24GB, used) — cheapest catalog machine that runs devstral 24B fully ($50)

H100 80GB PCIe (used) — upgrade path machine

every model the NVIDIA H100 NVL 94GB can run — full list

explore the full catalog90 machines indexed · live prices · what each one can run
2026-09-24 · ← all models on the NVIDIA H100 NVL 94GB · fit = weights + 4k KV vs usable memory