The Pulse
Google Makes Anthropic-Managed Claude Generally Available
Google Cloud’s Anthropic-managed Claude option gives customers access to Claude Platform features while Anthropic acts as the data processor. Google says the service is generally available as of October 23, 2026, alongside its existing Goog
AI.info Team · Written with AI · fact-checked ·
Google Cloud’s new Anthropic-managed Claude option puts a choice in front of customers: get Anthropic’s platform features as soon as they launch, or keep Google Cloud as the data processor and retain its governance controls. Google announced general availability on October 23, 2026, adding a second way to use Claude on its cloud infrastructure. Both options provide the latest Claude models, but the companies divide responsibility differently.
One cloud, two data processors
With the new deployment, Anthropic operates the service and processes data under Anthropic’s commercial terms, while Google Cloud provides the infrastructure. Customers use Google Cloud IAM for authentication, can keep Cloud Audit Logging, and pay through Google Cloud billing. Eligible Google Cloud spending commitments can cover Claude usage.
“Accessing new features on day one is important.”
Arvind Sundararajan, CTO, Abacus.AI
Google’s existing Google-managed option keeps Google Cloud as the data processor under customers’ existing agreements. It offers VPC Service Controls, customer-managed encryption keys and data residency in the U.S., EU and Asia-Pacific, along with integrations including BigQuery and Model Armor. Google says new Claude platform features reach that option after they are integrated into its infrastructure.
Anthropic’s tools arrive on day one
The Anthropic-managed service provides day-zero access not only to Claude models but also to Claude Platform features and betas. Google names Claude Managed Agents, which remains in beta, along with skills and code execution. Anthropic’s announcement also lists computer use, its Claude Console for managing and observing agents, and other platform capabilities.
The model lineup listed by Google includes Claude Fable 5.1, Opus 5.5, Sonnet 5.5 and Haiku 5.5. The companies say new models will be available through the Anthropic-managed option when they launch. Google-managed customers also get day-zero model access, but must wait for platform tools to be incorporated into Google’s version.
Billing stays with Google; processing does not
Customers can keep their Google Cloud identity and billing arrangements while using Anthropic’s service. Anthropic says the offering is available through U.S. and Global endpoints, and Claude usage can draw down an existing Google Cloud commitment. The arrangement does not make Google the data processor: Anthropic holds that role for the new option.
The distinction matters for organizations whose contracts or internal policies specify who handles prompts and other data. Google presents its own managed deployment as the fit for customers that need Google to process data, regional residency, or Google-native integrations. The Anthropic-managed route is aimed at teams that value immediate access to Anthropic’s full platform more than having Google serve as processor.
Availability comes with setup and discount caveats
Anthropic says customers can get started by opening the Anthropic-managed Google Cloud offering or consulting its documentation. The announcement says customers with existing discounts should contact their Google or Anthropic account executive before starting: discounts cannot be applied retroactively. The service is generally available, but customers still need to select the processor and deployment model that fits their requirements.
For Google Cloud users, the practical choice is now explicit: Anthropic-managed Claude brings the Claude Platform’s tools and betas on day one, while Google-managed Claude keeps Google as processor and retains Google’s stated residency and governance features.