Anthropic brings the full Claude Platform to Google Cloud

If you have ever tried to get enterprise approval for Claude, you know the drill: security wants existing cloud IAM, and finance demands you burn through your existing Google Cloud spend commitments first.

Anthropic just solved that headache. The company officially launched its Anthropic-managed Claude Platform on Google Cloud, giving enterprise customers identical features to native Claude without leaving their Google Cloud setup.

Native features with zero cloud lag

Typically, running third-party frontier models inside a hyperscaler means living weeks or months in the past. Features land on the native API first, while cloud marketplaces wait for integration updates.

Anthropic is eliminating that gap. With this new offering, Google Cloud customers get same-day access to new releases and betas the exact moment they launch on the native Claude Platform.

Teams can use their existing Google Cloud IAM credentials for permissions, roll all Claude usage into a single consolidated Google Cloud invoice, and count those dollars toward existing Google Cloud enterprise commitments.

Every major Claude capability is included out of the box: Claude Managed Agents, in-line code execution, Computer use, Files API, prompt caching, batch processing, and the MCP connector. Customers also get access to the Claude Console workspace.

Anthropic also made Fast Mode for Claude Opus 5.5 generally available on this setup. Fast Mode cranks output speed by up to 2.5x, priced at $8 per million input tokens and $40 per million output tokens for high-speed agent interactions.

Two distinct paths for cloud teams

Google Cloud customers now have two clear ways to deploy Claude.

This new Anthropic-managed path is built for developers who need day-zero feature parity, with Anthropic operating the underlying infrastructure and acting as the official data processor. It is currently available across US and Global endpoints.

Meanwhile, the existing Google-managed offering stays active for organizations that strictly require Google as their legal data processor, have rigid regional data residency rules, or rely on native Google protections like Model Armor and Agent Engine.

One logistical caveat: companies with existing negotiated discounts must contact their account executive before flipping the switch, as Anthropic noted discounts cannot be applied retroactively.

Why it matters

We think this is a masterclass in enterprise distribution.

For enterprise builders, it eliminates the friction between engineering ambition and corporate procurement. Developers no longer have to beg for separate API budgets or settle for delayed cloud ports; they can build with cutting-edge agent tools immediately while finance draws down existing cloud commits. If your organization already has Google Cloud commitments, moving your Claude workloads here is a no-brainer.

For the broader AI ecosystem, it challenges the traditional model reseller playbook. Anthropic gets access to Google enterprise billing pipelines without surrendering control over its feature roadmap or execution environment.

The real test will be whether enterprise risk teams sign off on Anthropic acting as the data processor within their Google tenant. If strict compliance forces high-security teams back into the Google-managed tier, Anthropic will have to manage a widening feature gap across two very different enterprise audiences.