Skip to main content
Version: Canary 🐤

Welcome to Anyscale

Bring your most ambitious AI workloads. Anyscale simplifies the infrastructure.

Build, deploy, and manage generative AI models and massive batch jobs, all on Anyscale's optimized compute platform.

Get started​

To start exploring, complete the following:

  • Sign up for an Anyscale account.
  • Check your email for a magic link.

Understand the basics​

Anyscale AI Platform​

Build modern machine learning systems and deploy them in a production-ready hosted environment. Scale sophisticated GenAI workloads like embedding, fine-tuning, batch inference, serving, and other distributed Ray workloads on dedicated compute.

Anyscale Endpoints​

Use this public API service to add an intelligence layer to any application with the best open source large language models. Fully hosted by Anyscale on shared compute resources, Anyscale Endpoints lets you scale fast without thinking about infrastructure.

How does it work?​

Anyscale hosted compute executes your code within its own dedicated cloud environment. Anyscale AI Platform workloads run on AWS or GCP within the United States.

Anyscale Endpoints uses resources shared between customers that may be hosted in one or more cloud regions set forth on the Cloud Services Provider list.

Our fully hosted compute model addresses complex infrastructure challenges on your behalf. You don't need to engage with Kubernetes, Docker, or manage an AWS account. You focus on the code, and Anyscale optimizes the compute.

Try it out​

Go to the console and start making API calls with an LLM example or launch an AI application template in Anyscale's fully hosted development environment.

Anyscale UI

Get support​

  • Official support: To reach out to the team directly, email preview-help@anyscale.com.
  • Community channel: Join the open community Slack to exchange ideas with fellow Anyscale users.
Special requests

Reach out to us at preview-help@anyscale.com if you want access to the following features:

  • Access to A100 and H100 GPUs
  • Fully hosted cloud on Google Cloud Platform
  • Custom compute with self-hosted cloud accounts
  • Custom cloud storage for data retention
Storage Retention Policy

Anyscale provides temporary object storage as scratch storage. $ANYSCALE_ARTIFACT_STORAGE retains data for up to 90 days from the date of creation. Anyscale removes all user data within a maximum of 120 days following account deactivation. Hosted Anyscale may retain logs for up to 2 years. Users can request deletion of their Hosted Anyscale account by contacting support@anyscale.com.