Configure Claude Cowork with Amazon Bedrock in 10 Minutes (No Anthropic Account Required)

This step‑by‑step guide shows how to set up Claude Cowork to use Amazon Bedrock as its inference backend—without registering an Anthropic account—by installing the desktop client, generating a Bedrock API key, enabling Developer Mode, configuring third‑party inference, editing the config file to unlock 1 M‑token context, and handling common issues on macOS and Windows.

Amazon Cloud Developers
Amazon Cloud Developers
Amazon Cloud Developers
Configure Claude Cowork with Amazon Bedrock in 10 Minutes (No Anthropic Account Required)

Overview

Claude Cowork is Anthropic’s AI‑assistant desktop application. By configuring Amazon Bedrock as the inference provider you can run Claude without an Anthropic account, keeping all model traffic and data within your own AWS account.

Prerequisites

A macOS or Windows computer.

An AWS account with Amazon Bedrock access enabled.

Bedrock access for Claude models in the target region (e.g., us-east-1).

Step 1 – Install Claude Desktop

Open a browser and go to https://claude.com/download. Download the .dmg for macOS or the .msix for Windows, install it, and launch the app. The login screen will show “Continue with Google” and “Enter your email”. Do not log in yet.

Step 2 – Generate an Amazon Bedrock API Key

In the AWS Management Console:

Search for “Amazon Bedrock” and open the service.

In the left navigation, choose Discover → API Keys .

Switch to the Long‑term API Keys tab and click Generate long‑term API key .

Give the key a name (AWS auto‑generates one like BedrockAPIKey‑xxxx‑at‑<your‑account‑id>), set an expiration (e.g., 365 days), and enable the “Permissions to access Amazon Bedrock Marketplace models” option. Click Generate and copy the key.

Step 3 – Enable Developer Mode in Claude

In Claude, go to the top menu Help → Troubleshooting → Enable Developer Mode . A new Developer menu appears.

Step 4 – Configure Third‑Party Inference

From the Developer menu choose Configure Third‑Party Inference… and fill in the fields:

Connection : Provider = Amazon Bedrock.

Bedrock Credentials : Paste the API key.

Identity & Models : In the Model list add the two model IDs (press Enter after each):

us.anthropic.claude-opus-4-7
us.anthropic.claude-sonnet-4-6

Leave Organization UUID and Credential helper script empty.

Click Apply locally . The configuration is written to a local JSON file and takes effect only on the current device.

Step 5 – Restart Claude and Select Bedrock

After applying, Claude prompts to relaunch. Click “Relaunch now”. On the login screen a new option Continue with Amazon Bedrock appears; select it to bypass Anthropic login.

Step 6 – Enable 1 M‑Token Context (Optional but Recommended)

Claude Desktop defaults to a 200 K context window. To unlock the 1 M context supported by Bedrock models you must edit the config file directly.

Open the config file:

macOS: open ~/Library/Application\ Support/Claude-3p/configLibrary/ Windows: explorer %APPDATA%\Claude-3p\configLibrary\ Locate the UUID‑named .json file and edit the inferenceModels field from a plain string array to an object array with "supports1m": true for each model. Example before and after:

"inferenceModels": ["us.anthropic.claude-opus-4-7","us.anthropic.claude-sonnet-4-6"]
"inferenceModels": [{"name":"us.anthropic.claude-opus-4-7","supports1m":true},{"name":"us.anthropic.claude-sonnet-4-6","supports1m":true}]

Save the file and fully quit Claude (e.g., killall Claude && open -a Claude on macOS or Cmd+Q on Windows), then restart.

After restarting, the model selector shows “(1M context)” variants for Opus 4.7 and Sonnet 4.6.

Security, Cost, and FAQ

API‑Key safety : Use short‑term keys (12 h) for production or rotate long‑term keys regularly.

Pricing : Bedrock usage is billed per‑consumption on your AWS bill; no extra Anthropic subscription is required. See the Bedrock pricing page for details.

Common Issues

Login still shows the normal page : Ensure Developer Mode is enabled, “Apply locally” succeeded, and Claude was fully restarted. Check logs at ~/Library/Logs/Claude/main.log (macOS) or %APPDATA%\Claude\logs\main.log (Windows).

Model errors after connection : Verify the AWS account has Claude model access in the selected region, the API key is valid, and model IDs include the us. prefix.

API‑Key security : Long‑term keys are for personal testing; production should use short‑term keys or IAM role authentication.

Cost calculation : Bedrock charges are consumption‑based; refer to the AWS Bedrock pricing page.

Windows‑Specific Deployment

Download the MSIX package from the same Claude download page and install with PowerShell: Add-AppxPackage -Path ".\Claude.msix" -Verbose Enable the Virtual Machine Platform feature (required for the sandbox):

Enable-WindowsOptionalFeature -Online -FeatureName VirtualMachinePlatform -All

After enabling, reboot the machine.

For headless or enterprise scenarios, configure Bedrock via the Windows registry. Create a .reg file (example below) and import it:

Windows Registry Editor Version 5.00

[HKEY_CURRENT_USER\Software\Policies\Claude]
"inferenceProvider"="bedrock"
"inferenceBedrockRegion"="us-east-1"
"inferenceBedrockBearerToken"="YOUR_BEDROCK_API_TOKEN"
"inferenceModels"="[{\"name\":\"us.anthropic.claude-opus-4-7\",\"supports1m\":true},{\"name\":\"us.anthropic.claude-sonnet-4-6\",\"supports1m\":true}]"

Replace YOUR_BEDROCK_API_TOKEN with the key generated in Step 2. Double‑click the .reg file and confirm the import.

After importing, launch Claude; it should start directly in inference mode with Cowork and Code tabs visible.

Quick Reference

For macOS and Windows, the guide includes screenshots (preserved below) illustrating each UI step, the API‑Key warning dialog, the configuration window, and the final 1 M‑context model selector.

Amazon Bedrock API Keys page
Amazon Bedrock API Keys page
Generate Long‑term API Key
Generate Long‑term API Key
Security risk confirmation
Security risk confirmation
Configuration file
Configuration file
1M context verification
1M context verification
Original Source

Signed-in readers can open the original source through BestHub's protected redirect.

Sign in to view source
Republication Notice

This article has been distilled and summarized from source material, then republished for learning and reference. If you believe it infringes your rights, please contactadmin@besthub.devand we will review it promptly.

ConfigurationAWSAI AssistantClaudeAmazon BedrockDeveloper Mode1M Context
Amazon Cloud Developers
Written by

Amazon Cloud Developers

Official technical community of Amazon Cloud. Shares practical AI/ML, big data, database, modern app development, IoT content, offers comprehensive learning resources, hosts regular developer events, and continuously empowers developers.

0 followers
Reader feedback

How this landed with the community

Sign in to like

Rate this article

Was this worth your time?

Sign in to rate
Discussion

0 Comments

Thoughtful readers leave field notes, pushback, and hard-won operational detail here.