Skip to main content
This guide shows you how to configure AWS Bedrock as an external model provider in AI Studio. After setting up this provider, you can use foundation models from AWS Bedrock when building agents.
AI Studio currently supports Bedrock models in us-east-1, us-west-1, and eu-west-1 regions only. See Regional considerations for details.

​
Prerequisites

Before adding Bedrock models to AI Studio, you need:
  • An AWS account with access to Amazon Bedrock
  • Model access enabled for the Bedrock models you want to use (see Enable model access)
  • IAM credentials with permission to invoke Bedrock models
AI Studio supports text generation and embedding models from Bedrock. Video, audio, and image generation models are not supported.

​
Configure AWS credentials

AI Studio supports two authentication methods for AWS Bedrock: access keys and IAM Role ARN. Choose the method that best fits your security requirements. Access keys provide a straightforward way to authenticate with AWS. Create a dedicated IAM user for AI Studio rather than using personal credentials.

​
Create an IAM user

  1. Sign in to the AWS IAM Console
  2. Navigate to Users in the left sidebar
  3. Select Create user
  4. Enter a username (for example, writer-aistudio-bedrock)
  5. Select Next
  6. Choose Attach policies directly
  7. Search for and select AmazonBedrockFullAccess, or create a custom policy with more restrictive permissions (see IAM policy requirements)
  8. Select Next, then Create user

​
Generate access keys

  1. Select the user you created
  2. Navigate to the Security credentials tab
  3. Under Access keys, select Create access key
  4. Choose Third-party service as the use case
  5. Acknowledge the recommendation and select Next
  6. Add an optional description tag (for example, AI Studio Bedrock integration)
  7. Select Create access key
  8. Copy the Access key ID and Secret access key
The secret access key is only shown once. Copy it immediately and store it securely. If you lose the secret key, you must create new access keys.
IAM roles provide more secure access without long-term credentials. With this method, you create a role that Writer’s AI Studio assumes to access your Bedrock models.

​
Create an IAM role

  1. Sign in to the AWS IAM Console
  2. Navigate to Roles in the left sidebar
  3. Select Create role
  4. Choose Custom trust policy
  5. Enter the following trust policy, which allows Writer to assume the role:
{
  "Version": "2012-10-17",
  "Statement": [
    {
      "Effect": "Allow",
      "Principal": {
        "AWS": "arn:aws:iam::WRITER_ACCOUNT_ID:root"
      },
      "Action": "sts:AssumeRole",
      "Condition": {
        "StringEquals": {
          "sts:ExternalId": "YOUR_EXTERNAL_ID"
        }
      }
    }
  ]
}
Contact Writer support to obtain the correct WRITER_ACCOUNT_ID and YOUR_EXTERNAL_ID values for your organization.
  1. Select Next
  2. Search for and select AmazonBedrockFullAccess, or attach a custom policy
  3. Select Next
  4. Enter a role name (for example, WriterAIStudioBedrockRole)
  5. Select Create role
  6. Copy the Role ARN from the role summary page (for example, arn:aws:iam::123456789012:role/WriterAIStudioBedrockRole)

​
IAM policy requirements

For production deployments, create a custom IAM policy with least-privilege permissions instead of using AmazonBedrockFullAccess. The minimum required permissions are: 
{
  "Version": "2012-10-17",
  "Statement": [
    {
      "Effect": "Allow",
      "Action": [
        "bedrock:InvokeModel",
        "bedrock:InvokeModelWithResponseStream"
      ],
      "Resource": [
        "arn:aws:bedrock:*::foundation-model/*"
      ]
    }
  ]
}
To restrict access to specific models or regions, modify the Resource field: 
{
  "Resource": [
    "arn:aws:bedrock:us-west-2::foundation-model/anthropic.claude-3-sonnet-*",
    "arn:aws:bedrock:us-west-2::foundation-model/meta.llama3-*"
  ]
}

​
Enable model access in Bedrock

Before using models in AI Studio, you must request access to them in the AWS Bedrock console.
  1. Navigate to the Amazon Bedrock console
  2. Select your preferred region from the region dropdown (for example, US West (Oregon))
  3. Navigate to Model access in the left sidebar
  4. Select Manage model access
  5. Check the boxes for the models you want to enable
  6. Select Request model access
  7. Wait for the status to change from In Progress to Access granted
Some models require you to submit a use case and accept additional terms. Follow the prompts in the console to complete the access request.
For the most current list of available models and regional availability, see the Amazon Bedrock model documentation.

​
Add Bedrock models in AI Studio

After configuring your AWS credentials and enabling model access:
  1. Navigate to Models & Guardrails > Models in AI Studio
  2. Select + Add model
  3. Select AWS Bedrock as the provider
  4. Choose your authentication method:
    • Access keys: Enter your access key ID and secret access key
    • Role ARN: Enter your IAM role ARN
  5. Select the AWS region where you enabled model access
  6. Select the models you want to add from the available list
  7. Configure team access:
    • All teams: Anyone with builder access can use the model
    • Specific teams: Restrict to selected teams
  8. Select Add model

​
Regional considerations

AI Studio supports Bedrock models in the following AWS regions:
  • US East (us-east-1)
  • US West (us-west-1)
  • EU West (eu-west-1)
Support for additional AWS regions is coming soon.
When configuring Bedrock in AI Studio:
  • Select a supported region where you have model access enabled
  • Ensure your IAM credentials have permissions in the same region you select
  • Some Bedrock models are only available in specific regions—check the Amazon Bedrock regional availability documentation
Ensure the region you select in AI Studio matches the region where you enabled model access in AWS. Your IAM user or role must have permissions in that same region.

​
Monitor costs

AWS bills Bedrock usage directly to your AWS account based on the tokens processed. AI Studio also tracks usage and costs for external models, providing visibility into spending across all your models in one place. For information about monitoring model health and automatic recovery, see Monitor model health.

​
Troubleshoot Bedrock configuration

​
Invalid credentials error

If you see an “Invalid credentials” or “Authentication failed” error:
  • Verify credentials are copied correctly without extra spaces
  • Check that access keys are still active in the IAM console
  • For Role ARN: Verify the trust policy includes the correct Writer account ID and external ID
  • Ensure the IAM user/role has the required Bedrock permissions

​
Model not available error

If a model doesn’t appear or returns an error:
  • Confirm you’ve requested and received access to the model in the Bedrock console
  • Verify you selected the correct region in AI Studio
  • Check that your IAM policy allows access to the specific model

​
Access denied error

If you see permission errors when using a model:
  • Review your IAM policy to ensure it includes bedrock:InvokeModel permission
  • Check the policy’s Resource field allows access to the model you’re using
  • For Role ARN: Verify the role can be assumed by Writer’s account

​
Unhealthy model status

If a model shows as unhealthy in AI Studio:
  • AI Studio automatically retries unhealthy models after a cooldown period
  • For transient issues like temporary AWS outages, no action is needed
  • For persistent issues, check the troubleshooting items above

​
Next steps