Icon for SaladCloud

SaladCloud

Container GPU cloud on a distributed network of consumer RTX GPUs, billed per second by priority tier

SaladCloud runs containers on a distributed network of consumer RTX GPUs, from home PCs to workstation fleets. Its Salad Container Engine schedules container groups across nodes and moves a workload to another node of the same class when one disconnects. vCPU and RAM are included with every GPU.

Pricing is per GPU-hour, billed per second, in four priority tiers: an RTX 4090 costs $0.16/hr at Lowest and $0.33 at High, which is never preempted by other workloads. Salad itself rules out multi-GPU NVLink training, sub-10 ms latency and strict data residency.

Pricing: Hourly

Hosting Cloud
Pricing $0.015/GPU/hr (older GPU classes, Lowest priority)
HQ ๐Ÿ‡บ๐Ÿ‡ธ United States
Screenshot of SaladCloud webpage

Work on SaladCloud? Feature it at the top of Inference APIs.

Is your product missing?

Add it here →