Anthropic said on Thursday that Claude Models on Google Cloud (Anthropic managed) is now generally available, giving Google Cloud customers the complete Claude developer platform with the same billing, credentials and spending commitments they already use on Google Cloud. New platform features and betas ship on the Anthropic-managed offering the same day they go live on Anthropic’s own platform, while Anthropic operates the service and acts as the data processor.
The launch makes Google Cloud the only hyperscaler where all three closed-model ecosystems — Anthropic’s Claude, Google’s own Gemini and OpenAI’s models, which arrived on Google Cloud’s Model Garden a day earlier — can be bought against existing cloud commitments. Google announced the general availability on its Cloud blog on October 22.
What Google Cloud Customers Get
Customers using the Anthropic-managed option log in with their existing Google Cloud IAM credentials, pay through a single Google Cloud invoice and burn down their existing Google Cloud commitments with Claude usage. The offering ships with US and Global endpoints, and Anthropic says new models will land on it as they launch. Available models include Claude Fable 5.1, Opus 5.5, Sonnet 5.5 and Haiku 5.5.
The feature list reads like the full Claude platform catalogue: Claude Managed Agents (beta) for long-running agents without managing infrastructure, code execution for running Python and building visualisations inside an API call, computer use for operating a browser or desktop, agent skills for repeatable workflows, an advisor strategy (beta) that consults an advisor model, plus the Files API, citations, web search and web fetch, an MCP connector (beta), prompt caching and batch processing. Fast Mode on Claude Opus 5.5 is also generally available on the offering, promising up to 2.5x faster output token generation at $8 per million input tokens and $40 per million output tokens.
Customers also get access to the Claude Console, Anthropic’s development environment for writing prompts, running agents on hosted infrastructure, monitoring usage and costs and managing access per workspace. Anthropic adds one administrative note: customers with existing discounts should contact their Anthropic or Google Cloud account executive before starting, because discounts cannot be applied retroactively.
Two Deployment Options, Two Data Processors
Google Cloud now sells two versions of Claude side by side, and the difference comes down to who processes the data and how early new features arrive. In the Anthropic-managed option, Anthropic operates the stack, acts as the data processor and ships day-zero access to every beta and generally available platform feature. In the Google-managed option, Google Cloud is the data processor, a choice Anthropic says suits customers with strict regional data-residency requirements or who want native Google integrations such as Model Armor or Agent Engine.
Both options deliver the same frontier models on day zero, the same pricing and the same Google Cloud billing, according to Google. The split mirrors how enterprises actually buy: some want the newest agent tooling the moment it exists — the same tooling behind products like Anthropic’s proactive Claude Duet agent app — while others need their existing compliance posture untouched.
Why It Matters
The Google Cloud catalogue has become the neutral shelf of the AI industry. Within 48 hours, the cloud division announced general availability for Anthropic’s full platform and private previews of OpenAI’s GPT-6 models, both payable against the commitments enterprises already hold. For buyers, that turns model choice into a procurement decision rather than a contract negotiation: the same spend commitment can now reach Claude, GPT or Gemini, and the differentiators become governance tooling, data-processor terms and how fast new features arrive. The risk is concentration — every model bought through Google Cloud deepens the customer’s dependence on that cloud — but the direction of travel is clear: the hyperscalers are competing to be the place where every model is sold, including their owners’ direct rivals.
Background
Anthropic has distributed Claude through Google Cloud for years via the Google-managed Vertex AI route, but developers on that path waited for new platform features to be integrated into Google’s infrastructure. The Anthropic-managed offering removes that lag by letting Anthropic run the service itself. The timing is pointed: OpenAI’s models arrived on Google Cloud’s Model Garden the day before, and both announcements land as Anthropic prepares a public stock offering, with a public S-1 expected as early as this week.
Source: Anthropic
0 Comments