What's New

AWS has launched Claude Platform on AWS in general availability, making it the first cloud provider to offer direct access to Anthropic's native Claude Platform experience through an existing AWS account. The service eliminates the need for separate Anthropic accounts, billing, or credential management by integrating directly into the AWS ecosystem. Customers gain access to Anthropic's full suite of platform capabilities—including beta features—while retaining familiar AWS security and billing tooling.

How It Works

• The service is operated by Anthropic, meaning customer data is processed outside the AWS security boundary, but access is provisioned and authenticated through standard AWS IAM credentials and access controls.
• Developers interact with Anthropic's native Claude Platform APIs and Console directly, gaining access to the same interface and feature set available on Anthropic's own platform.
• AWS CloudTrail audit logging is enabled for the service, providing organizations with a full audit trail of API calls and access events within their existing AWS security monitoring workflows.
• Billing is consolidated into the customer's existing AWS account, removing the need to manage a separate Anthropic billing relationship or payment method.
• Available features include Claude Managed Agents (beta), advisor strategy (beta), web search, web fetch, code execution, Files API (beta), Skills (beta), MCP connector (beta), prompt caching, citations, batch processing, and the Claude Console for prompt development and evaluation.
• The MCP (Model Context Protocol) connector enables integration with external tools and data sources, while the Files API and Skills features support richer, stateful agentic workflows.

Why It's Important

• Organizations that have standardized on AWS can now access Anthropic's cutting-edge and beta-stage AI capabilities without fragmenting their vendor relationships, procurement processes, or security posture.
• Consolidated billing simplifies cost attribution and financial governance for teams already managing AI/ML spend through AWS Cost Explorer and budgeting tools.
• CloudTrail integration means security and compliance teams get unified audit visibility across both AWS-native services and Claude Platform activity, reducing blind spots in regulated environments.
• Access to early-access beta features—such as Managed Agents, Skills, and the MCP connector—gives AWS customers a competitive edge by enabling experimentation with Anthropic's latest agentic and tool-use capabilities before they reach general availability.
• The service lowers the operational overhead for development teams by removing the need to manage separate API keys, account provisioning, and usage tracking for Anthropic's platform.

How It's Different

• Unlike Amazon Bedrock, which hosts Claude models within AWS's own infrastructure and security boundary, Claude Platform on AWS is operated by Anthropic and processes data outside the AWS security boundary—offering Anthropic's native experience rather than a managed AWS-hosted version.
• AWS is the first and currently only cloud provider to offer this native Claude Platform integration, giving AWS customers exclusive early access to this distribution model.
• Customers get access to Anthropic's beta and early-access features (e.g., Managed Agents, Skills, MCP connector) that are not yet available through Amazon Bedrock's Claude integration.
• The Claude Console for prompt development and evaluation is included, providing a richer developer experience compared to purely API-based access through Bedrock.
• Rather than requiring a net-new vendor relationship with Anthropic, the entire onboarding, authentication, and billing flow is handled through the customer's existing AWS account.

When to Prefer It

• Choose Claude Platform on AWS when your team wants access to Anthropic's latest beta features and agentic capabilities (e.g., Managed Agents, MCP connector, Skills) that are not yet available in Amazon Bedrock.
• Prefer this service when your organization wants to consolidate AI vendor billing and access management under a single AWS account without maintaining a separate Anthropic account.
• Use it when your development workflow benefits from the Claude Console's native prompt development and evaluation environment, rather than building custom tooling.
• This is the right choice when your use case does not have strict regional data residency or data sovereignty requirements, since data is processed outside the AWS security boundary by Anthropic.
• Ideal for enterprises and development teams that are already deeply invested in AWS IAM and CloudTrail for security governance and want to extend that coverage to their Anthropic API usage.
• Consider it when you need rapid access to Anthropic's full platform roadmap, including features that may reach Anthropic's platform before they are integrated into Bedrock.

Availability

• Status: Generally Available (GA) as of May 11, 2026; several features (Managed Agents, advisor strategy, Files API, Skills, MCP connector) remain in beta.
• Supported Regions: US East (N. Virginia), US East (Ohio), US West (Oregon), Canada (Central), South America (São Paulo), Europe (Dublin, London, Frankfurt, Milan, Zurich, Paris, Stockholm), Asia Pacific (Tokyo, Seoul, Sydney, Melbourne, Jakarta).
• Data Boundary: Customer data is processed by Anthropic outside the AWS security boundary; not suitable for workloads with strict data residency or sovereignty requirements.
• Pricing: Pricing details were not disclosed in the announcement; customers should refer to the Claude Platform on AWS product page for current pricing, as it is likely consumption-based and billed through AWS.
• Prerequisites: Requires an existing AWS account; access is managed via IAM credentials with no separate Anthropic account needed.
• Limitations: Not a replacement for Amazon Bedrock for workloads requiring AWS-native data residency, VPC integration, or compliance with AWS security boundary guarantees.