OCPI·LIVE
H100from $1.95H100from $1.66H200from $0.50B200from $3.74A100from $0.47L40Sfrom $0.60RTX 4090from $0.29RTX 5090from $0.40H100from $1.95H100from $1.66H200from $0.50B200from $3.74A100from $0.47L40Sfrom $0.60RTX 4090from $0.29RTX 5090from $0.40
gpuprice.fyi
MethodologyData statusAPI
EN中
Index / Compare
Cloud GPU Price Comparison

A100 vs RTX 5090

A100 80GB vs RTX 5090 — current cloud rental prices normalized to USD per GPU per hour, across providers. RTX 5090 is currently the cheaper option — about 76% less than A100 80GB ($0.40 vs $1.67 per GPU per hour).

A100 80GB
NVIDIA Ampere · UNKNOWN
$1.67/GPU/hr
index median · 2 providers
range $0.47–3.28 · 5 providers
cheapest: Vast.ai $0.47
↗ Full A100 80GB prices
RTX 5090
NVIDIA Blackwell · PCIE
$0.40/GPU/hr
cheapest listing (no index median yet)
range $0.40–2.00 · 3 providers
cheapest: Vast.ai $0.40
↗ Full RTX 5090 prices
“

Listed price ≠ transaction price

These are published on-demand list prices, normalized per GPU per hour. The index median uses reliable-tier (datacenter) providers only. Read the methodology →