What is Vertex AI?
Vertex AI is Google Cloud's platform for training, deploying and calling machine-learning and foundation models. In April 2026 Google folded it into the Gemini Enterprise Agent Platform ("Agent Platform" for short), but the name survives in URLs, SDK clients and pricing pages, so most people still say Vertex AI. Claude runs on it as a partner model: you enable it in a Google Cloud project, authenticate with Google Cloud IAM, and call an API that is nearly identical to the Claude Messages API. Google operates the inference and handles the data; usage is billed to the Google Cloud account at Google's prices.
When should you use Vertex AI for Claude?
It is the natural path for GCP-native teams and for companies that need EU data residency — Google's eu multi-region endpoint keeps inference inside Europe, at a premium over global routing. Provisioned throughput and FedRAMP High via Assured Workloads are further reasons. The trade-offs: among Anthropic's server-side tools only web search is available (no code execution, Files, Batches, Skills or Managed Agents), and model lifecycle dates are set by Google. For plain inference and tool use it is a solid choice; we build on it in AI & Automation engagements on Google Cloud. All seven ways to get Claude are compared in Seven paths to Claude.