Claude gets an Anthropic-run option on Google Cloud
A new deployment choice lets Google Cloud customers run the full Claude Platform under Google billing and IAM, with new features landing the same day as the direct API.
Anthropic and Google Cloud made “Claude Models on Google Cloud (Anthropic managed)” generally available this week. It sits alongside the existing Google-managed option on Vertex AI, but Anthropic runs the infrastructure directly, acting as the data processor under its own terms rather than Google’s.
The payoff is feature parity: Managed Agents (beta), Skills, code execution, computer use and the Files API all show up on day zero, the same day they ship on Anthropic’s own platform. The Google-managed route integrates those features over time instead. You still authenticate with Google Cloud IAM, get Cloud Audit Logging, and billing counts toward your existing Google Cloud spend commitments. Both options offer Claude Fable 5.1, Sonnet 5.5, Haiku 5.5 and Opus 5.5 at the same price.
Why it matters: teams that consolidated onto Google Cloud for billing reasons no longer have to wait for Anthropic’s newest capabilities to reach Vertex AI. If you’ve been running Claude through Vertex and missing something newly launched on the direct API, this is the fix, with no separate contract or bill.
The caveat: the Google-managed option still exists for a reason. It offers VPC Service Controls, CMEK and regional data residency across the US, EU and APAC. The Anthropic-managed option doesn’t mention any of those, so pick based on whether compliance controls or feature speed matter more to you.