xpu liveBETA
GPU comparisonGPU providers
GPU explorer / Providers / SaladCloud
S

GPU CLOUD PROVIDER

SaladCloud

Distributed GPU cloud

Public website GPU-hour prices at four priorities. Optional organization API via SALAD_USE_API=1, SALAD_API_KEY and SALAD_ORGANIZATION; account quotes stay private.

Compare 12 offers ↓Provider website ↗

RENTAL OPTIONS

GPU offers 12

USD / GPU-hour

12 offers

NVIDIA RTX 3090Community / batchServerless
$0.09USD / GPU-hour$0.09 / GPU-hour
GPU time only

Public USD per GPU-hour; billed per second while running. High priority avoids priority preemption but nodes may disconnect.

Minimum 1 GPU(s)Region Confirm with provider
Server configuration
CPU / vCPU
Not published for this offer; confirm with provider
System RAM
Not published for this offer; confirm with provider
Storage
Not published for this offer; confirm with provider
Network / egress
Not published for this offer; confirm with provider
GPU allocation
Not published for this offer; confirm with provider
Multi-GPU topology
Not published for this offer; confirm with provider

Server resources are specific to this offer and are separate from the GPU’s own VRAM and hardware capabilities.

Listed · Auto-collected · Checked 2026-10-03View provider pricing ↗
NVIDIA RTX 3090Community / highServerless
$0.17USD / GPU-hour$0.17 / GPU-hour
GPU time only

Public USD per GPU-hour; billed per second while running. High priority avoids priority preemption but nodes may disconnect.

Minimum 1 GPU(s)Region Confirm with provider
Server configuration
CPU / vCPU
Not published for this offer; confirm with provider
System RAM
Not published for this offer; confirm with provider
Storage
Not published for this offer; confirm with provider
Network / egress
Not published for this offer; confirm with provider
GPU allocation
Not published for this offer; confirm with provider
Multi-GPU topology
Not published for this offer; confirm with provider

Server resources are specific to this offer and are separate from the GPU’s own VRAM and hardware capabilities.

Listed · Auto-collected · Checked 2026-10-03View provider pricing ↗
NVIDIA RTX 3090Community / lowServerless
$0.1170USD / GPU-hour$0.117 / GPU-hour
GPU time only

Public USD per GPU-hour; billed per second while running. High priority avoids priority preemption but nodes may disconnect.

Minimum 1 GPU(s)Region Confirm with provider
Server configuration
CPU / vCPU
Not published for this offer; confirm with provider
System RAM
Not published for this offer; confirm with provider
Storage
Not published for this offer; confirm with provider
Network / egress
Not published for this offer; confirm with provider
GPU allocation
Not published for this offer; confirm with provider
Multi-GPU topology
Not published for this offer; confirm with provider

Server resources are specific to this offer and are separate from the GPU’s own VRAM and hardware capabilities.

Listed · Auto-collected · Checked 2026-10-03View provider pricing ↗
NVIDIA RTX 3090Community / mediumServerless
$0.1430USD / GPU-hour$0.143 / GPU-hour
GPU time only

Public USD per GPU-hour; billed per second while running. High priority avoids priority preemption but nodes may disconnect.

Minimum 1 GPU(s)Region Confirm with provider
Server configuration
CPU / vCPU
Not published for this offer; confirm with provider
System RAM
Not published for this offer; confirm with provider
Storage
Not published for this offer; confirm with provider
Network / egress
Not published for this offer; confirm with provider
GPU allocation
Not published for this offer; confirm with provider
Multi-GPU topology
Not published for this offer; confirm with provider

Server resources are specific to this offer and are separate from the GPU’s own VRAM and hardware capabilities.

Listed · Auto-collected · Checked 2026-10-03View provider pricing ↗
NVIDIA RTX 4090Community / batchServerless
$0.16USD / GPU-hour$0.16 / GPU-hour
GPU time only

Public USD per GPU-hour; billed per second while running. High priority avoids priority preemption but nodes may disconnect.

Minimum 1 GPU(s)Region Confirm with provider
Server configuration
CPU / vCPU
Not published for this offer; confirm with provider
System RAM
Not published for this offer; confirm with provider
Storage
Not published for this offer; confirm with provider
Network / egress
Not published for this offer; confirm with provider
GPU allocation
Not published for this offer; confirm with provider
Multi-GPU topology
Not published for this offer; confirm with provider

Server resources are specific to this offer and are separate from the GPU’s own VRAM and hardware capabilities.

Listed · Auto-collected · Checked 2026-10-03View provider pricing ↗
NVIDIA RTX 4090Community / highServerless
$0.33USD / GPU-hour$0.33 / GPU-hour
GPU time only

Public USD per GPU-hour; billed per second while running. High priority avoids priority preemption but nodes may disconnect.

Minimum 1 GPU(s)Region Confirm with provider
Server configuration
CPU / vCPU
Not published for this offer; confirm with provider
System RAM
Not published for this offer; confirm with provider
Storage
Not published for this offer; confirm with provider
Network / egress
Not published for this offer; confirm with provider
GPU allocation
Not published for this offer; confirm with provider
Multi-GPU topology
Not published for this offer; confirm with provider

Server resources are specific to this offer and are separate from the GPU’s own VRAM and hardware capabilities.

Listed · Auto-collected · Checked 2026-10-03View provider pricing ↗
NVIDIA RTX 4090Community / lowServerless
$0.2170USD / GPU-hour$0.217 / GPU-hour
GPU time only

Public USD per GPU-hour; billed per second while running. High priority avoids priority preemption but nodes may disconnect.

Minimum 1 GPU(s)Region Confirm with provider
Server configuration
CPU / vCPU
Not published for this offer; confirm with provider
System RAM
Not published for this offer; confirm with provider
Storage
Not published for this offer; confirm with provider
Network / egress
Not published for this offer; confirm with provider
GPU allocation
Not published for this offer; confirm with provider
Multi-GPU topology
Not published for this offer; confirm with provider

Server resources are specific to this offer and are separate from the GPU’s own VRAM and hardware capabilities.

Listed · Auto-collected · Checked 2026-10-03View provider pricing ↗
NVIDIA RTX 4090Community / mediumServerless
$0.2730USD / GPU-hour$0.273 / GPU-hour
GPU time only

Public USD per GPU-hour; billed per second while running. High priority avoids priority preemption but nodes may disconnect.

Minimum 1 GPU(s)Region Confirm with provider
Server configuration
CPU / vCPU
Not published for this offer; confirm with provider
System RAM
Not published for this offer; confirm with provider
Storage
Not published for this offer; confirm with provider
Network / egress
Not published for this offer; confirm with provider
GPU allocation
Not published for this offer; confirm with provider
Multi-GPU topology
Not published for this offer; confirm with provider

Server resources are specific to this offer and are separate from the GPU’s own VRAM and hardware capabilities.

Listed · Auto-collected · Checked 2026-10-03View provider pricing ↗
NVIDIA RTX 5090Community / batchServerless
$0.25USD / GPU-hour$0.25 / GPU-hour
GPU time only

Public USD per GPU-hour; billed per second while running. High priority avoids priority preemption but nodes may disconnect.

Minimum 1 GPU(s)Region Confirm with provider
Server configuration
CPU / vCPU
Not published for this offer; confirm with provider
System RAM
Not published for this offer; confirm with provider
Storage
Not published for this offer; confirm with provider
Network / egress
Not published for this offer; confirm with provider
GPU allocation
Not published for this offer; confirm with provider
Multi-GPU topology
Not published for this offer; confirm with provider

Server resources are specific to this offer and are separate from the GPU’s own VRAM and hardware capabilities.

Listed · Auto-collected · Checked 2026-10-03View provider pricing ↗
NVIDIA RTX 5090Community / highServerless
$0.50USD / GPU-hour$0.5 / GPU-hour
GPU time only

Public USD per GPU-hour; billed per second while running. High priority avoids priority preemption but nodes may disconnect.

Minimum 1 GPU(s)Region Confirm with provider
Server configuration
CPU / vCPU
Not published for this offer; confirm with provider
System RAM
Not published for this offer; confirm with provider
Storage
Not published for this offer; confirm with provider
Network / egress
Not published for this offer; confirm with provider
GPU allocation
Not published for this offer; confirm with provider
Multi-GPU topology
Not published for this offer; confirm with provider

Server resources are specific to this offer and are separate from the GPU’s own VRAM and hardware capabilities.

Listed · Auto-collected · Checked 2026-10-03View provider pricing ↗
NVIDIA RTX 5090Community / lowServerless
$0.3330USD / GPU-hour$0.333 / GPU-hour
GPU time only

Public USD per GPU-hour; billed per second while running. High priority avoids priority preemption but nodes may disconnect.

Minimum 1 GPU(s)Region Confirm with provider
Server configuration
CPU / vCPU
Not published for this offer; confirm with provider
System RAM
Not published for this offer; confirm with provider
Storage
Not published for this offer; confirm with provider
Network / egress
Not published for this offer; confirm with provider
GPU allocation
Not published for this offer; confirm with provider
Multi-GPU topology
Not published for this offer; confirm with provider

Server resources are specific to this offer and are separate from the GPU’s own VRAM and hardware capabilities.

Listed · Auto-collected · Checked 2026-10-03View provider pricing ↗
NVIDIA RTX 5090Community / mediumServerless
$0.4170USD / GPU-hour$0.417 / GPU-hour
GPU time only

Public USD per GPU-hour; billed per second while running. High priority avoids priority preemption but nodes may disconnect.

Minimum 1 GPU(s)Region Confirm with provider
Server configuration
CPU / vCPU
Not published for this offer; confirm with provider
System RAM
Not published for this offer; confirm with provider
Storage
Not published for this offer; confirm with provider
Network / egress
Not published for this offer; confirm with provider
GPU allocation
Not published for this offer; confirm with provider
Multi-GPU topology
Not published for this offer; confirm with provider

Server resources are specific to this offer and are separate from the GPU’s own VRAM and hardware capabilities.

Listed · Auto-collected · Checked 2026-10-03View provider pricing ↗

Prices are dated listings, not live availability. Confirm the exact GPU edition and total charges with the provider. Storage, networking and taxes may be extra.