Introduction to Claude Cowork and Amazon Bedrock
Today marks an exciting development in AI technology as Claude Cowork becomes available within Amazon Bedrock. This integration allows users to leverage the powerful capabilities of Claude Code Desktop through Amazon Bedrock, either directly or via an LLM gateway.
Organizations ranging from startups to global enterprises across various industries are already building with Claude Code in Amazon Bedrock to enhance developer productivity and expedite delivery. Amazon Bedrock facilitates this by allowing users to build within their existing AWS environments, ensuring enterprise security, regional data residency, and scalable inference. Importantly, your data remains under your control, as Amazon Bedrock does not store prompts, files, tool inputs and outputs, or model responses, nor does it use them for training base models.
Now, with Claude Cowork in Amazon Bedrock, AI adoption can be extended to all knowledge workers in an organization. This desktop application can read documents, perform multi-step searches, process files, and return completed work, significantly enhancing productivity.
Understanding Claude Cowork
Claude Cowork allows users to delegate tasks such as search, document analysis, data processing, and report generation to Claude from a desktop application. Users benefit from Claude Desktop’s core features, including projects, artifacts, memory, file upload and export, remote connectors, skills, plugins, and MCP servers. However, features that require Anthropic-hosted inference, such as Chat Tab, Computer Usage, and Skills Marketplace, are excluded because Claude Cowork exclusively routes model inference through Amazon Bedrock in your AWS account. For a complete feature comparison with Claude Enterprise, see Features on 3P.
Pricing for Claude Cowork is based on consumption through your existing AWS agreement and billing, without needing an Anthropic HQ license.
Integration of Claude Cowork with Amazon Bedrock
Amazon Bedrock serves as the inference backend in your AWS account and supported AWS Regions.
Setting up Claude Cowork in Amazon Bedrock involves two main steps. Initially, users download the Claude Desktop application onto their machines. Subsequently, your device management system (such as Jamf, Microsoft Intune, or Group Policy) provides a configuration to Claude Desktop that enables inference mode. This configuration specifies the Amazon Bedrock model ID and inference profile, authentication method, and organizational policies. If your organization centralizes access to models through an LLM gateway, you can point Claude Desktop to the gateway URL through the same managed configuration.
Organizations already utilizing Claude Code in Amazon Bedrock can employ the same configuration for Claude Cowork.
Figure 1: The following diagram illustrates the end-to-end flow
The application features three outgoing paths, all controllable by you. Model inference is transmitted to Amazon Bedrock in the AWS Regions you configure. Connections to the MCP server, if set up, access endpoints you trust. Anthropic only receives aggregate telemetry (token count, model ID, error codes, anonymous device ID), which can be disabled via configuration options.
Amazon Bedrock offers intra-region, cross-geography, and global cross-region inference profiles, enabling you to choose the appropriate level of data residency for your organization.
Claude Cowork integrates seamlessly with the AWS services you already use.
For more details on MDM configuration, credentials, MCP servers, and plugins, see the Claude Cowork configuration reference.
Claude Cowork in Practical Application
Once integrated, users can open Claude Desktop and begin delegating tasks. Claude Cowork connects to external data sources through MCP servers, allowing access to live documentation, web searches, and other tools during operation.
For instance, consider a product manager contemplating a new notification feature for a college athletics application hosted on AWS. With client meeting notes pointing in different directions and a set of project requirements, the manager faces a time crunch. They upload these to Cowork.
Claude then compares disparate inputs and synthesizes them into a comprehensive product brief. It evaluates the proposed approach, explores alternatives, identifies technical challenges, and supports recommendations with evidence. By connecting to the AWS Documentation MCP server and an MCP web search server, Claude bases its file on current service documentation, market context, and competitor positioning.
Figure 2: The product manager transforms meeting notes into a product brief with Claude Cowork
Within minutes, the product manager has a structured brief based on current sources ready for review. Other knowledge workers can also benefit from this approach. An operations manager can consolidate scattered documentation into an SOP, a financial analyst can transform raw data into a formatted monthly review, and a research team can compile results from multiple sources into a single report.
Conclusion
Claude Cowork in Amazon Bedrock enables extensive AI adoption within organizations while keeping data secure within your AWS environment. Available on macOS and Windows in AWS Regions where Claude templates are available, it offers a seamless experience. To get started, download Claude Desktop from claude.com/download and review the Claude Cowork Setup Guide.
Try Claude Cowork today and share your feedback through AWS re:Post for Amazon Bedrock or via your usual AWS Support contacts.
About the Authors
Sofiane Hamiti
Sofiane Hamiti is a technology leader with over 12 years of experience in building AI solutions and leading high-performing teams to maximize client outcomes. He is passionate about empowering diverse talents to make a global impact and achieve their professional aspirations.
Ryan Ray
Ryan Ray is a Senior Partner Solutions Architect and AI Technical Lead at AWS, serving as the Global Technology Lead for Anthropic at AWS. He works at the intersection of cloud architecture and artificial intelligence, helping organizations adopt and scale Anthropic technologies on AWS.
Antonio Rodriguez
Antonio Rodriguez is a Senior Specialty Solutions Architect for Amazon Bedrock at AWS, specializing in enterprise generative AI architecture and regulated industrial deployments.
For more information, visit the source link Here.
