#1
Vast.aiMarketplace
Configuration1× RTX 4090 · 12 vCPU
RegionEurope
PurchaseOn-demand
$0.32/ GPU / hr
NVIDIA · Ada Lovelace
RTX 409024 GB GDDR6X
A cost-focused option for smaller inference jobs, image generation, and experimentation.
Lowest observed rate$0.32per GPU / hour · sample feedView RTX 4090 on Flatkey Compute ↗
ArchitectureAda Lovelace
VRAM24 GB
Memory bandwidth1.01 TB/s
Published2022
Price board
2 configurations tracked
Median sample rate: $0.45 per GPU-hour.
#2
RunpodGPU cloud
Configuration1× RTX 4090 · 16 vCPU
RegionGlobal
PurchaseOn-demand
$0.59/ GPU / hr