Verda raises prices for the ninth time — all nine GPUs up about 1% again

What changed, read on verda.com/pricing on Nov 4, 2026 (unrounded page figures in brackets): GB300 $13.18 → $13.31 [$13.31], B300 $11.46 → $11.57 [$11.57], B200 $8.96 → $9.05 [$9.048], H200 $6.31 → $6.38 [$6.376], H100 SXM $4.84 → $4.89 [$4.889], A100 80GB $2.27 → $2.29 [$2.293], A100 40GB $1.63 → $1.64 [$1.641], RTX PRO 6000 $2.67 → $2.70 [$2.698] and L40S $1.92 → $1.94 [$1.939]. Each move is within a whisker of +1.0% — a uniform list adjustment, not one card repricing.

The pattern: Verda rose on Oct 23, 24, 26, 27, 28, 29 and 30, held on Oct 31 and Nov 1, rose an eighth time on Nov 2, held on Nov 3 — and has now risen a ninth time. Its GB300, the only GB300 rate we list, has gone from $11.82 an hour at our Oct 19 check to $13.31 today, up 12.6% in just over two weeks. Verda is not alone in the pattern: Nebius raised across the board on Oct 1, RunPod Secure on Oct 14, SaladCloud's Priority rise took effect Nov 1 and RunPod Serverless rose by Nov 2. Rising list prices are consistent with growing demand (our reading, labelled as interpretation; capacity and cost moves can produce the same shape).

What it costs: Verda holds the board's Best offer on only two cards — the GB300 and the A100 40GB — and both rose today, so those two best offers rose with it. Everywhere else the cheapest verified offers are other providers' and none moved: the board's H100 SXM best is still TensorDock at $2.25 an hour (at an 8-GPU minimum) and its B200 best is still RunPod Community at $5.98. If you rent at Verda around the clock, today's rise adds about $95 a month per GB300, $66 per B200 and $37 per H100 on a 730-hour month — small per hour, real per fleet.

Checked today and unchanged: RunPod Secure and RunPod Serverless, Lambda, Nebius, CoreWeave, Thunder Compute and TensorDock all held at the rates read on Nov 3, and SaladCloud's Priority rates (RTX 5090 High $0.60, RTX 4090 High $0.44) are still the live ones. Not checked today, and named rather than covered by the stamp: RunPod Community (its tier could not be separated in today's read), Hyperstack (its pricing table could not be read, for the seventh straight day) and nafnaf.ai (its homepage showed no GPUs available today, so there was no live listing to verify).