Amazon Bedrock brings Claude Opus 5, Sonnet 5 and Haiku 4.5 to India

How India geographic inference works

Amazon Bedrock offers cross-Region inference profiles, which distribute inference across multiple AWS Regions so customers do not have to manage capacity in each one. A request originates in the source Region where the API call is made and is automatically routed to one of the destination Regions defined in the profile.

According to AWS, the India geographic profile keeps inference within India. Requests route only between ap-south-1 and ap-south-2, the Mumbai and Hyderabad Regions. Input prompts and output results might move between those two Regions. The India option is offered in addition to the already supported global cross-Region inference, and AWS says it can help customers who need to process data locally in a desired geography.

Because requests draw on a broader pool of compute rather than the capacity of a single Region, AWS says this helps maintain throughput and consistent performance under load, which matters most during traffic peaks.

Data handling, billing and supported features

AWS states that cross-Region inference runs over the secure AWS network with end-to-end encryption for data in transit. Customer data is not stored in a destination Region and remains exclusively within the source Region.

Amazon Bedrock uses a zero data retention (ZDR) data security model, which means that by default it does not store model inputs or outputs. AWS adds that certain models require human review by AWS if content is flagged by automatic safety classifiers, and points to the Data retention section of the Amazon Bedrock User Guide for details.

Billing and quota consumption are tracked against the customer's account in the source Region, regardless of which backend Region handled the request. Amazon CloudWatch and AWS CloudTrail record log entries in the source Region only, so monitoring stays in one place.

Geographic cross-Region inference is available on the bedrock-runtime endpoint. It supports Anthropic's Messages API and the native Amazon Bedrock InvokeModel and Converse APIs, along with features such as Amazon Bedrock Guardrails and intelligent prompt routing.

Getting started in the console and in code

In the Amazon Bedrock console, users can try the models in the text playground without any coding or SDK setup. The steps described by AWS are:

• Open the console in the Region to be used as the source.
• Under Test, choose Playground, then Select model.
• Search for Anthropic Claude Opus 5 and select IN Anthropic Claude Opus 5 as the inference profile under Inference.
• Choose Apply, enter a prompt and choose Run.

For programmatic access, developers use the India geographic inference profile ID. The Anthropic Messages API works through the Anthropic SDK, while the InvokeModel and Converse APIs work through the AWS CLI and AWS SDK. AWS's examples use ap-south-1 as the Region and profile IDs such as in.anthropic.claude-sonnet-5 and in.anthropic.claude-opus-5.

Prerequisites include an active AWS account with Amazon Bedrock access, the AWS CLI installed and configured, Python 3.8+, Boto3, the Anthropic SDK, and the Bedrock Token Generator for model inference authentication. The IAM role or user needs permission to invoke models using a geographic cross-Region inference profile. Usage, performance and costs can be monitored through CloudWatch and AWS Cost Explorer.

What it means for customers in India

For organizations in India that need to keep inference within the country, the launch adds Claude Opus 5, Claude Sonnet 5 and Claude Haiku 4.5 as options on Amazon Bedrock. AWS says customers can build scalable, resilient generative AI applications while keeping inference within the country. For the latest model availability by Region, AWS refers users to the Regional availability by models page in the Amazon Bedrock User Guide.