Anthropic has expanded the availability of its Claude AI models in India with the introduction of in-country inference through Amazon Bedrock, enabling prompts and model responses from Indian customers to be processed on servers located within the country.
The move is aimed at addressing the growing need for data residency and local processing capabilities, particularly among enterprises and organisations operating in regulated sectors. Through the India endpoint on Amazon Bedrock, businesses can use Claude for sensitive AI workloads while keeping the processing of those requests within India.
Claude Opus 5, Claude Sonnet 5 and Claude Haiku 4.5 are now available in India through Amazon Bedrock's geographic cross-Region inference. The setup allows requests to be routed between AWS's Mumbai and Hyderabad regions while keeping inference within the country's geographical boundaries. AWS said the approach is designed to provide access to a broader pool of computing capacity while maintaining local processing.
The development comes as Anthropic continues to see strong adoption of Claude in the Indian market. India is currently the company's second-largest market for Claude.ai, with software development accounting for 45.2% of work-related tasks undertaken by users in the country, according to Irina Ghose, Managing Director, Anthropic India.
The use of Claude in India has also expanded beyond software development, with organisations across banking and payments, IT services, healthcare, science, digital businesses and public services adopting the technology. TCS is equipping 50,000 associates with Claude, while Infosys is integrating the model with its Topaz platform. Cognizant is using Claude in systems developed for its clients, while L&T Technology Services is incorporating the technology into its engineering platforms.
For businesses handling sensitive information, the India inference capability could be particularly significant as data handling, regulatory compliance and residency requirements become increasingly important considerations in enterprise AI deployment.
Ghose said the local processing capability was driven by requests from customers in India, particularly those in regulated industries, who wanted to use Claude for critical workloads while ensuring that every request was processed within the country. She added that financial institutions, government organisations and large enterprises had sought in-country inference to meet their regulatory and data-handling requirements.
Amazon Bedrock provides the infrastructure for accessing Claude locally, while AWS's geographic cross-Region inference profile ensures that requests remain within India's Mumbai and Hyderabad regions. AWS also states that customer data is not stored in a destination region during cross-Region inference and that Amazon Bedrock operates under a zero-data-retention model by default for model inputs and outputs.
The launch could also help unlock AI deployments that were previously held back by data residency considerations, giving Indian enterprises a pathway to integrate advanced generative AI capabilities into workloads where local processing is a prerequisite.
With enterprise adoption of generative AI accelerating across sectors, Anthropic's India expansion through Amazon Bedrock marks another step in its efforts to deepen its presence in one of the world's fastest-growing AI markets, while giving organisations greater flexibility to deploy Claude within local data-processing requirements.
