Select Language:
If you’re working with newer Bedrock models in certain regions, like eu-west-1, you might find that you can’t invoke these models directly using their model IDs. Instead, AWS requires you to use an inference profile. Here’s a simple guide to help you get started and make the process smooth.
First, you’ll need to find the inference profiles available in your region. You can do this by running the following command:
bash
aws bedrock list-inference-profiles –region eu-west-1 \
–query “inferenceProfileSummaries[?contains(inferenceProfileName, ‘Sonnet’)].[inferenceProfileId,inferenceProfileName]” \
–output table
This command will list available profiles, showing entries like:
eu.anthropic.claude-sonnet-4-20250514-v1:0— which is the EU regional profile.global.anthropic.claude-sonnet-4-20250514-v1:0— the global cross-region profile.
Once you’ve identified the correct inference profile, you should use its ID instead of the model ID when invoking the model. Here’s how you can do that:
python
response = bedrock.invoke_model(
modelId=”eu.anthropic.claude-sonnet-4-20250514-v1:0″, # inference profile ID, not a model ID
contentType=”application/json”,
accept=”application/json”,
body=json.dumps({…})
)
Keep in mind, your IAM policy needs to include permissions for the inference profile. The resource ARNs are a bit different from standard model ARNs. For inference profiles, the ARN format looks like this:
json
{
“Effect”: “Allow”,
“Action”: “bedrock:InvokeModel”,
“Resource”: [
“arn:aws:bedrock:::foundation-model/anthropic.“,
“arn:aws:bedrock:eu-west-1:ACCOUNT_ID:inference-profile/eu.anthropic.*”
]
}
Notice that the inference profile ARN includes your specific AWS account ID and region, unlike regular foundation model ARNs.
The reason behind this setup is that AWS manages regional routing and capacity using inference profiles. For example, a profile starting with eu. routes requests to EU-based infrastructure, while a global. profile can handle cross-region requests. Directly using the model ID only works where on-demand throughput is supported explicitly.
By following these steps, you’ll be able to smoothly invoke newer Bedrock models in regions where direct model ID access isn’t available.





