SaladCloud
Distributed GPU cloud that runs containers on shared consumer GPUs at lower cost.
Introduction
SaladCloud runs containerized AI and batch workloads on a network of consumer GPUs contributed by their owners. Users choose GPU types and quantities, and the Salad Container Engine handles orchestration and reallocation when a machine goes offline. The platform also offers a transcription API with a free trial for new organizations.
- Added on
- Nov 2, 2026
- Website
- salad.com
- Pricing model
- Free Trial
- Company
- Salad Technologies
Product information
What is SaladCloud?
SaladCloud is a distributed GPU cloud that deploys containerized workloads across a network of shared GPUs.
How to use SaladCloud?
Containerize your model or inference server, select GPU types and quantity, and deploy it with the Salad Container Engine.
Core features of SaladCloud
- Distributed GPU cloud
- Lower-cost GPU pricing
- Scalable compute resources
- Container Engine for orchestration
- Transcription API
- Gateway service and virtual kubelets
Use cases of SaladCloud
- 1Image generation
- 2Voice AI and speech workloads
- 3Computer vision
- 4Batch processing
- 5Language model inference
Frequently asked questions about SaladCloud
How does SaladCloud work?+
Workloads run on shared GPUs similar to spot instances, and the platform manages orchestration and reallocation.
Do I pay for cold start time?+
According to the site, billing starts only after the container is running.
Does SaladCloud have a free trial?+
The Container Engine has no trial, but the Transcription API includes a free trial of five audio hours for new organizations.
Which category is SaladCloud in?+
SaladCloud is listed under AI Code Editor, AI Transcription.