RunPod Serverless logo

RunPod Serverless

Cloud GPU for auto-scaling Machine Learning

Core Product

Containers Edge Compute Docker
Free Tier Free Trial

Information Resources

RunPod Serverless Homepage
RunPod Serverless Homepage

What is RunPod Serverless?

RunPod, founded in 2022, is a cloud GPU computing platform that specializes in AI (Artificial Intelligence) and ML (Machine Learning) workloads. The company positions itself as a developer’s launchpad for full-stack AI applications, providing the essential computational backbone that allows businesses to run AI and ML workloads efficiently, globally, and at scale.

Headquartered in San Francisco, RunPod operates with a remote and hybrid team structure that spans the US, Canada, and Europe. In December 2023, RunPod successfully raised $20 million in a Seed VC funding round, led by significant investors including Intel Capital and Dell Technologies Capital.

RunPod Serverless Focus

RunPod specializes in providing cloud GPU and CPU computing services for developing, training, and scaling AI applications. The platform offers:

  • A globally distributed GPU cloud
  • Template environments for quick setup
  • Serverless endpoints for deployment
  • Docker images that can be deployed as a scalable Serverless Workers

RunPod Serverless Key Features

  • Instant Deployment: RunPod has reduced cold-boot times to milliseconds, allowing users to start building within seconds of deploying pods.
  • 50+ Ready-to-Use Templates: Users can choose from over 50 pre-configured environments like PyTorch and TensorFlow, or bring their own custom containers.
  • Extensive CPU Options: Ranging from Intel 11th gen i5 to 13th gen i9 and AMD 8 core EPYC to 64 core Ryzen Threadripper, and more.
  • Flexible GPU Options: RunPod offers various GPU types, including NVIDIA H100s, A100s, and options to reserve AMD MI300Xs and AMD MI250s in advance.
  • Development Tools: The platform provides tools for seamless development, including instant hot-reloading for local changes and easy testing through CLI-provided endpoints.

Features

Service Types

Containers
Edge Compute

Functions Runtimes

Python

Container Runtimes

Docker

Execution Limits and Resource Quotas

Min. Memory
8GB
Max. Memory
192GB
Default Timeout
10 min
Max. Timeout
> 5s
Request Payload
10MB / 20MB
Response Payload
10MB / 20MB

Compliance Certificates

GDPR

Regional Availability

Deployment Scope
Global / Regional
Regions
NA, EU
Points of Presence
9, 5
Last modified: September 19, 2024

*Information is subject to change. Please consult the official RunPod Serverless website for more details.