NVIDIA RTX 4090

Cloud GPU Pricing — Live from 8+ Providers
Starting From
$0.160/hr
5 on-demand listings
VRAM
24 GB GDDR6X
Architecture
Ada Lovelace
FP16 Performance
165 TFLOPS
Memory Bandwidth
1.0 TB/s
Listings
7 options
Provider Pricing
ProviderPrice / hrEst. MonthlyTypeVRAMAction
BestSalad Cloud $0.1600/hr $116.8/mo On-Demand 24 GB Rent Now →
FluidStack $0.3200/hr $233.6/mo On-Demand 24 GB Rent Now →
TensorDock $0.3770/hr $275.21/mo On-Demand 24 GB Rent Now →
Vast.ai $0.6696/hr $488.81/mo On-Demand 24 GB Rent Now →
RunPod $0.6900/hr $503.7/mo On-Demand 24 GB Rent Now →
Vast.ai $0.1950/hr $142.35/mo Spot 24 GB Rent Now →
RunPod $0.3400/hr $248.2/mo Spot 24 GB Rent Now →
About the NVIDIA RTX 4090

The NVIDIA RTX 4090 is the best consumer GPU for AI inference, offering exceptional price-to-performance for smaller models. With 24 GB GDDR6X it can run 13B parameter models at full precision and 70B models with quantization. The go-to for Stable Diffusion, real-time inference, and budget-conscious ML workloads.

GPU Specifications
VRAM24 GB GDDR6X
ArchitectureAda Lovelace
FP16 Performance165 TFLOPS
Memory Bandwidth1.0 TB/s
TDP450W
InterconnectPCIe 4.0
Best For
LLM InferenceStable DiffusionFine-tuning Small ModelsImage Generation

Compare All NVIDIA RTX 4090 Options Side by Side

Filter by price type, sort by cost, and find the best deal for your workload.

Open Full Comparison →