Home

Free Tools

Salad - GPU Cloud

Affordable, scalable distributed GPU cloud designed for AI and machine learning workloads.

Visit Site

AI Developer Tools AI API AI Transcription Large Language Models (LLMs)AI Text-to-Speech

About Salad - GPU Cloud

SaladCloud is a distributed GPU cloud platform that enables users to reduce cloud costs by up to 90% compared to hyperscalers. Access over 60,000 active GPUs daily, starting at just $0.02 per hour. Ideal for AI inference, batch processing, molecular dynamics, and more, SaladCloud leverages unused global compute resources to create a fully people-powered cloud environment.

How to Use

Deploy AI/ML models at scale securely on SaladCloud by containerizing your applications, selecting desired resources, and letting SaladCloud handle deployment. You can also earn rewards by sharing your unused GPU hours, contributing to a sustainable compute network.

Features

Secure Gateway Service

Advanced Transcription API

Distributed GPU Cloud Platform

Elastic Compute Resources

Container Management Engine

Secure and Reliable Deployment

Affordable GPU Pricing Options

Virtual Kubernetes (Kubelets)

Use Cases

Molecular Dynamics Simulations

Training Language Models

Computer Vision Applications

Voice AI Development

Large-Scale Batch Processing

Data Collection and Analysis

AI Image Generation

Best For

Data scientistsAI development teamsStartups scaling AIAcademic researchersML engineersSoftware developers

Pros

Cost savings of up to 90% compared to hyperscale providers

Simple deployment with Salad Container Engine

Utilizes unused GPU resources sustainably

Access to a vast network of GPUs worldwide

No pre-paid contracts with flexible pricing

Secure platform with SOC2 certification

Cons

Potential interruptions due to compute sharing

Maximum VRAM limited to 24 GB

Longer startup times than traditional clouds

Not suitable for ultra-low latency workloads

Pricing Plans

Choose the perfect plan for your needs. All plans include 24/7 support and regular updates.

RTX 5090

$0.294 per hour (Batch)

32GB VRAM, 8GB Memory, 4 vCPUs

Get Started

RTX 5080

$0.219 per hour (Batch)

16GB VRAM, 8GB Memory, 4 vCPUs

Get Started

RTX 4090

$0.204 per hour (Batch)

24GB VRAM, 8GB Memory, 4 vCPUs

Get Started

RTX 3090 Ti

$0.154 per hour (Batch)

24GB VRAM, 8GB Memory, 4 vCPUs

Get Started

RTX 3090

$0.124 per hour (Batch)

24GB VRAM, 8GB Memory, 4 vCPUs

Get Started

RTX 4080

$0.154 per hour (Batch)

16GB VRAM, 8GB Memory, 4 vCPUs

Get Started

RTX 4070 Ti

$0.124 per hour (Batch)

12GB VRAM, 8GB Memory, 4 vCPUs

Get Started

RTX 3080 Ti

$0.124 per hour (Batch)

12GB VRAM, 8GB Memory, 4 vCPUs

Get Started

RTX 3060

$0.084 per hour (Batch)

12GB VRAM, 8GB Memory, 4 vCPUs

Get Started

RTX 3080

$0.114 per hour (Batch)

10GB VRAM, 8GB Memory, 4 vCPUs

Get Started

RTX 3070 Ti

$0.094 per hour (Batch)

8GB VRAM, 8GB Memory, 4 vCPUs

Get Started

RTX 3070

$0.094 per hour (Batch)

8GB VRAM, 8GB Memory, 4 vCPUs

Get Started

RTX 3060 Ti

$0.064 per hour (Batch)

8GB VRAM, 8GB Memory, 4 vCPUs

Get Started

General Purpose Instance

$0.005 per hour

1GB Memory, 1 vCPU

Get Started

CPU-Optimized Instance

$0.006 per hour

2GB Memory, 1 vCPU

Get Started

Memory-Optimized Instance

$0.012 per hour

8GB Memory, 1 vCPU

Get Started

Frequently Asked Questions

Find answers to common questions about Salad - GPU Cloud

What types of GPUs are available on SaladCloud?

All GPUs are Nvidia RTX or GTX series, selected for high-performance AI computing capabilities. Only top-tier, AI-enabled GPUs are onboarded.

How is security maintained on SaladCloud?

We encrypt containers during transit and at rest. Containers run in isolated environments, ensuring data security and consistent compute environments across nodes.

What are the key features of SaladCloud?

SaladCloud operates as a compute-sharing network with longer cold start times and potential interruptions. Max VRAM is 24 GB, making it unsuitable for extremely low-latency tasks.

What is Salad Container Engine (SCE)?

SCE simplifies deploying workloads via Docker containers. Containerize your models, select hardware, and let SaladCloud handle orchestration and deployment.

How does SaladCloud ensure workload availability?

SaladCloud uses performance ratings and proprietary tests to select optimal GPUs. When a GPU goes offline, workloads are automatically reallocated to similar resources.

Why do owners share their GPUs with SaladCloud?

Owners earn rewards, such as Salad tokens, by sharing idle compute resources. Many earn $30 to $200 monthly, which they can redeem for various rewards.

What if someone tries to access my container?

Our security systems detect unauthorized access attempts. If detected, we automatically terminate the environment and blacklist the host machine.

How do I know if a GPU is reliable and online?

We use a trust rating system to evaluate GPU performance and availability. The platform automatically reallocates workloads when resources go offline.

Do I pay for cold start or initialization times?

No, you are only billed for actual usage once your container is active. Cold start times are free of charge.

Are volume discounts available?

Yes, discounts are available for deployments of 50 or more GPU instances. Contact us through the demo form for customized pricing.