Claude Platform on AWS adds full Anthropic API access
Anthropic said the Claude Platform on AWS is now generally available, giving AWS customers access to the full Claude API feature set with AWS authentication, billing, and commitment retirement. The service includes Claude Managed Agents, code execution, web search, Files API, Skills, and other features that were previously tied to the native Claude API.
Anthropic said the Claude Platform on AWS is now generally available, giving AWS customers access to the full set of Claude API features through AWS accounts and billing.
The company said the service brings the complete Claude Platform to AWS customers for the first time, with new features and beta releases shipping the same day they appear on the native Claude API. Anthropic said customers use AWS IAM for authentication, CloudTrail for audit logging, and a single AWS invoice for billing and commitment retirement.
⚡ New to this?
This matters because it gives AWS customers a direct path to Anthropic’s full Claude API feature set without leaving AWS billing and identity tools. AWS IAM is the system companies use to control who can access what, and CloudTrail is AWS’s logging tool for tracking activity.
For non-experts, the big change is that teams can now build AI agents, run code, and connect to documents and web data in one place. That can reduce integration work, but it also raises the importance of knowing where data is processed and which platform model fits a company’s compliance needs.
🦞 OpenClaw angle
If you build self-hosted automation on AWS, compare this launch against your current Bedrock setup before moving workloads. Claude Platform on AWS gives you managed agents, code execution, Files API, and MCP connectors, so it may replace custom glue code around tool calls and document handling.
If your workflow depends on strict AWS-bound data processing, stay on Bedrock; if you want the full Claude feature set, test the new platform with a non-critical agent first. Also review IAM policies and CloudTrail coverage now, because the service plugs into your existing AWS identity and audit setup rather than creating a separate control plane.
According to Anthropic, teams can use their existing AWS credentials and IAM policies, which means they can work within the permissions and controls they already manage. The company said Claude Platform on AWS will be available in most AWS commercial regions and will support global and U.S. inference geographies.
The launch includes a long list of platform features. Anthropic said customers can use Claude Managed Agents, which is in beta, to build and deploy agents at scale. The platform also includes advisor strategy, a beta feature that has agents consult an advisor model, plus web search and web fetch for pulling in current information from the internet.
Other included features are code execution for running Python code, generating visualizations, and analyzing data directly inside API calls. Anthropic also listed the Files API, currently in beta, for uploading and referencing documents across conversations, and Skills, also in beta, for teaching Claude best practices so it produces more consistent results.
The company said the platform includes an MCP connector beta that can connect Claude to remote MCP servers without client-side code. MCP, or Model Context Protocol, is used to connect AI systems to tools and data sources. Anthropic also said prompt caching is included to reduce cost and latency on repeated context, along with citations for grounding answers in source documents and batch processing for high-volume asynchronous jobs.
Customers on the Claude Platform on AWS also get access to the Claude Console, Anthropic’s development environment for building and testing with Claude. The Console includes management tools for agents, skills, environments, vaults, and observability.
Anthropic said Claude Opus 4.7, Sonnet 4.6, and Haiku 4.5 are available on the platform, and new models will ship there when they launch.
Anthropic contrasted the new service with Claude on Amazon Bedrock, which remains available. The company said the difference is who operates the service and where data is processed. Claude Platform on AWS is operated by Anthropic and processes data outside the AWS boundary, while Amazon Bedrock keeps AWS as the data processor and processes data inside AWS’s infrastructure.
According to Anthropic, the new offering is intended for companies that want the full Claude Platform experience, while Bedrock is a better fit for organizations that need strict regional data residency or require data to be processed exclusively inside AWS.
Anthropic said the Claude Platform on AWS is available today. The company also said customers with an existing Bedrock private offer should contact their Anthropic or AWS account executive before getting started so discounts are applied correctly, since discounts cannot be applied retroactively to usage incurred before a Claude Platform private offer is accepted.
