Anyscale is listed with paid pricing.

What compliance standards does Anyscale support?

Anyscale is marked as compliant with SOC 2 Type II, ISO/IEC 27001, GDPR (DPA available).

Anyscale

Paid

Anyscale is a scalable AI compute platform built on Ray that allows teams to train, fine-tune, and serve large AI models at scale with flexible cloud infrastructure and managed Ray Serve endpoints.

Category

Coding

Subcategory

API Development

Free Tier

Free credits for new accounts

Paid Plans

Pay-as-you-go, Enterprise custom

API Cost

Endpoint API pricing available

Web AppAPICLI

Anyscale Trust & Security Report

Certifications, AI-training posture, and security controls, with sourced proof.

VIEW

●SOC 2 Type II●ISO/IEC 27001●GDPR (DPA available)

Verified against Anyscale's own trust and security pages. See the full report for sourced proof →

Region

Dual-plane architecture: the Customer Data Plane (where customer data and workloads run) executes inside the customer's own cloud account or Kubernetes cluster (AWS, GCP, Azure) via Bring-Your-Own-Cloud, so sensitive data stays in the customer's VPC. Only orchestration metadata flows to the Anyscale Control Plane. Application logs and customer data are stored locally on the cluster; ingested logs for the log viewer are opt-in by default for Anyscale Hosted clouds. Subprocessors are US-based (AWS, Google, Microsoft, Pylon) with Tailscale in Canada.

Trains on Data

Self-hostable

Yes

Training large AI models at scaleDistributed model fine-tuning with RayServing LLMs in productionScaling AI workloads across clustersCost-effective GPU compute for AI research