Google Cloud adds Anthropic-managed Claude deployment
Google Cloud's Anthropic-managed Claude deployment option is now generally available, sitting beside the Google-managed option that prioritises governance controls.
What Google announced
On October 23, 2026, Google Cloud said the Anthropic-managed deployment option for Claude on Google Cloud is now generally available. It runs beside the Google-managed option the company already offered, so customers pick who processes their data and how quickly they receive new Claude platform features. The detail is in Google Cloud's post on expanding Claude availability and deployment choice.
Both routes serve the same models from day one, at the same pricing, with billing handled through Google Cloud. The models the post names are Claude Fable 5.1, Sonnet 5.5, Haiku 5.5 and Opus 5.5, reachable through Google Cloud IAM and Marketplace billing.
How the two options differ
With the Anthropic-managed option, Anthropic operates the stack on Google Cloud infrastructure and acts as the data processor under its own commercial terms. Customers still authenticate with Google Cloud IAM, keep Cloud Audit Logging, and can draw the spend from eligible Google Cloud commitments. The advantage claimed here is launch-day access to Claude Platform features, specifically Claude Managed Agents, which the post marks as beta, along with skills and code execution.
With the Google-managed option, Google Cloud is the data processor under existing customer agreements. The post lists VPC Service Controls, Customer-Managed Encryption Keys and data residency in the U.S., EU and APAC. It also points to native integrations with BigQuery and Model Armor. New Claude Platform features arrive on this option later, once they are integrated into the Google-managed infrastructure.
Since model availability is day-zero on both, the real differences are the processing contract and how fast new tooling shows up.
What changes for people building apps
If a feature you depend on ships as a Claude Platform capability, the Anthropic-managed route is the one that has it immediately. Code execution is the clearest case in the post: an app that needs the model to run code as part of a request would otherwise wait for that capability to be integrated on the Google-managed side.
If your requirements are about data handling, the Google-managed route is where the post places the encryption keys, network controls and regional residency. Those controls are not described for the Anthropic-managed option, so a team with residency obligations has one clear path in this announcement.
Limits and open questions
The post names three platform features but does not list the full set, so it is hard to judge how much of a head start the Anthropic-managed option really gives. There is no timeline for when Claude Managed Agents, skills or code execution reach the Google-managed option. Residency and encryption details appear only for the Google-managed side, and the Anthropic-managed side gets no region list in the post. Pricing is described as identical and standard, with no rates published. The post also carries an October 23, 2026 date while the announcement circulated a day earlier.
What this suggests
This likely signals that Google is treating Anthropic's platform layer, not just its models, as something worth hosting, and that it expects customers to choose based on compliance posture rather than model access. It also suggests that teams wanting the newest Claude tooling will, for now, accept Anthropic's terms as processor. Whether that split holds as features get integrated is not answered.
Building on top of a model choice
The deployment option behind an API is usually invisible to the app calling it, which matters when you are prototyping. Draftship builds an app from a plain-language description, with a built-in database, sign-in and hosting, and paid plans add a choice of AI model and source code access (see Draftship pricing). If you are deciding which model an app you generate should call, the guide to using the AI chat and choosing a model covers that step. Draftship is not affiliated with Google Cloud or Anthropic.
Primary sources
- Expanding Claude availability and deployment choice on Google Cloud · Google Cloud AI and ML, Oct 22, 2026
This article was drafted with AI assistance from the primary sources listed below and checked automatically against them. An editor hasn't reviewed it yet.