AI compute market signals and learning

GPU Pricing

Current H200 Cloud Pricing

Sourced H200 on-demand rates across providers — list prices, H100-equivalent rates, and dated public price-page links.

H200 on-demand capacity is sourced from 5 provider price pages, each linked and dated below. The H100e column normalizes every rate to an H100-equivalent hour, so the H200 premium reads against the H100 table on one axis.

$3.50–$10 (5 providers)H200 on-demand band

Lowest to highest sourced public list price, as of Jun 11, 2026

5Providers sourced

Public price pages with an H200 on-demand rate that cleared editorial review

$2.40–$12.29 (7 providers)H100 band, for contrast

The prior-generation read the H200 premium is usually priced against

Sourced rates

Per-provider H200 on-demand rates

5 sourced provider rows, lowest list price first. Every row links the public price page it was observed on.

Sourced on-demand H200 GPU-hour rates
ProviderRegionList $/GPU-hrH100e $/hrAvailabilityNo change inSource
HyperstackMulti-region$3.50$3.04Available1 dayOpen price page
CrusoeMulti-region$4.29$3.73Available0 daysOpen price page
RunPod SecureMulti-region$4.39$3.82Available0 daysOpen price page
CoreWeaveus$6.30$5.48Available1 dayOpen price page
Oracleus-ashburn-1$10.00$8.70Available Limited26 daysOpen price page

Public on-demand list prices from provider price pages — not negotiated quotes, reserved terms, or spot rates. Each row carries its own source link and price date. How we source, label, and date evidence: methodology.

See every reviewed provider row

How to read it

What the H200 premium buys

H200 is the same Hopper generation as H100 with the memory roughly doubled — 141 GB of HBM3e against 80 GB. The premium over an H100 rate is paying for memory headroom, not extra raw compute.

When the premium is worth it

Memory-bound work — long contexts, large KV caches, bigger batches per GPU — gets more from the 141 GB than from a faster chip. Compute-bound training often does not.

Read it through the H100e column

The H100e $/hr column normalizes each H200 rate to an H100-equivalent hour, so a cheaper-looking H200 row and an H100 row compare on the same axis.

Generation vs generation

Where H200 sits between H100 and B200 on memory, performance per dollar, and availability — and which generation fits which workload.