Updated 4 days ago
Claude adds an Anthropic-managed Google Cloud route with different processor and feature choices

Claude adds an Anthropic-managed Google Cloud route with different processor and feature choices

The new service combines Claude Platform features with Google Cloud IAM and billing. Teams still need to choose the right processor, endpoints and integrations.

A second Google Cloud route gives access to the Claude Platform

Claude Models on Google Cloud (Anthropic managed) is generally available, giving Google Cloud customers a route to the Claude Platform through their existing Google Cloud authentication and billing. Anthropic announced the offering on October 22 and says it includes the full set of Claude API features, with new features and betas available on the same day as the native Claude Platform. [The official announcement describes the service](https://claude.com/resources/articles/claude‑models‑on‑google‑cloud‑anthropic‑managed). The important choice is which managed offering a team uses. Claude remains available through the Google‑managed service as well. The two routes differ in who operates and processes the service, the available endpoints and the integrations a team needs. Selecting a Claude model on Google Cloud is therefore not enough to identify those boundaries; the managed offering matters too.

Google Cloud handles access and billing; Anthropic operates the service

The Anthropic‑managed route works with Google Cloud IAM credentials. Billing goes through a single Google Cloud invoice, and Anthropic says Claude usage draws down an existing Google Cloud commitment. Anthropic operates the service and acts as the data processor. These are [separate roles in the announced arrangement](https://claude.com/resources/articles/claude‑models‑on‑google‑cloud‑anthropic‑managed): a Google Cloud access and billing layer does not mean Google is the processor for this offering. For a team deciding whether to adopt it, that separates two practical reviews. Existing IAM administration and cloud purchasing arrangements may fit the access and billing path, while the processing arrangement still needs to match the team’s requirements. The announcement describes the service design; OpenTools has not tested an organization’s permissions, invoice treatment or contractual setup.

Choose the Google‑managed route for different processor and regional needs

The Anthropic‑managed service currently offers US and Global endpoints. Anthropic says the Google‑managed offering remains the appropriate option for customers who need Google Cloud as the data processor, have strict regional data residency requirements, or want native integrations such as Model Armor or Agent Engine. [The release explicitly preserves both paths](https://claude.com/resources/articles/claude‑models‑on‑google‑cloud‑anthropic‑managed). That makes the processor and endpoint questions useful early filters. A team should establish which processor, geographic arrangement and Google‑native integrations its workload requires before treating feature access as the only deciding factor. The presence of a Global endpoint does not, by itself, establish a specific regional residency guarantee. The announcement does not present the Anthropic‑managed service as replacing every reason to choose the Google‑managed one.

Platform parity includes features that remain in beta

The offering includes code execution, computer use and [Agent Skills](https://opentools.ai/news/anthropic‑agent‑skills‑aaif‑donation‑maintainers‑membership), along with the Files API, citations, web search, web fetch, prompt caching and batch processing. Anthropic also lists Claude Managed Agents, advisor strategy and the MCP connector with their beta labels. The promise of same‑day access to native Platform features is [Anthropic’s stated availability policy](https://claude.com/resources/articles/claude‑models‑on‑google‑cloud‑anthropic‑managed), rather than an interoperability or performance test by OpenTools. General availability of the managed offering does not turn each included beta into a generally available feature. Teams can use the feature list to identify the API capabilities their application needs, then check those capabilities’ own documentation and release status. Anthropic also says the offering provides access to the Claude Console for configuring and observing agents, monitoring usage and costs, and managing workspace access and security.

Confirm existing discounts before generating usage

Customers with an existing discount should contact their Anthropic or Google Cloud account executive before starting with the Anthropic‑managed offering, according to the announcement. Anthropic says discounts cannot be applied retroactively to usage. [That instruction is part of the launch notice](https://claude.com/resources/articles/claude‑models‑on‑google‑cloud‑anthropic‑managed), so a shared invoice or commitment drawdown should not be taken as confirmation that a particular negotiated discount is already attached. The [Anthropic‑managed service documentation](https://platform.claude.com/docs/en/build‑with‑claude/claude‑models‑on‑google‑cloud/anthropic‑managed) and the [Google Cloud Marketplace listing](https://console.cloud.google.com/marketplace/product/anthropic/anthropic‑managed‑claude) are the setup destinations linked by Anthropic. For builders, the release expands the combination of Claude Platform features and Google Cloud access arrangements. The complete selection still depends on the processor, endpoints, required integrations, feature status and commercial terms for the workload.

Share this article

PostShare

Related News