Anthropic brings Claude in-country inference to India via Amazon Bedrock

Anthropic has launched in-country inference for its Claude models in India through Amazon Bedrock, allowing requests sent through the India endpoint to be processed within the country. The move is aimed at helping regulated institutions and large enterprises deploy Claude at scale while meeting data residency and governance requirements.

The capability covers Claude models including Claude Opus 5, Sonnet 5, and Haiku 4.5. Anthropic said the move will enable Indian financial institutions, government bodies and large enterprises operating under stringent regulatory and data-handling requirements to use Claude for sensitive workloads while keeping inference within Indian borders.

Data residency key to enterprise AI adoption

India is among Anthropic’s largest markets globally and has one of the company’s most technically sophisticated developer communities. According to Anthropic, Indian organisations have consistently sought the ability to run Claude inference locally, particularly as data residency becomes an important requirement for moving AI deployments from pilots to production.

“In August, we committed to bringing Claude inference to India, and we are delivering on that promise. Our customers, especially in regulated sectors and government, want to use Claude on their most critical work, with requests processed in the country. From today, they can,” said Irina Ghose, Managing Director, India, Anthropic. “This is a long-term commitment. We will keep growing our team in India and investing in local partnerships.”

Sandeep Dutta, President, AWS India and South Asia, said responsible AI adoption at scale had become a strategic priority for some of India’s most regulated industries. “For them, data residency is the parameter that decides whether AI stays confined to pilots or reaches full production. With in-country inference for Claude now available through Amazon Bedrock, requests are processed on servers based in India, enabling banks, insurers, and public institutions to trust and deploy AI at a production scale,” Dutta said.

Claude deployments through Amazon Bedrock in India also provide audit trails and access controls required by risk and compliance teams before AI systems can be deployed in production environments.

Reliance, CRED among early customers

Anthropic said Reliance, CRED and other organisations tested the capability during a private preview. Early customers and partners cited lower latency, performance, resilience and regulatory alignment as key benefits of processing Claude requests locally.

“As Anthropic’s customer and partner in India, we are pleased with this announcement. Local inference gives us the performance, resilience and regulatory alignment we need to move Claude into production areas across the Reliance ecosystem,” Reliance Industries said.

Amit Kapur, Chief AI & Services Transformation Officer at TCS, said demand for enterprise-grade AI deployments was growing in regulated industries. “Anthropic's in-country inference capability in India is an important step in meeting evolving data residency and governance requirements. Together, we are helping enterprises accelerate their journey,” Kapur said.

Vijay Narayanan, Head of Innovation & AI at Kotak Bank, said in-country inference would expand the options available to Indian banks. “In-country inference widens the options available to Indian banks. At Kotak, we built a single AI platform with governance designed in from the start, which lets us adopt new capabilities like this responsibly and at pace,” Narayanan said.

For Mahindra Group, the launch comes as data sovereignty gains importance under India’s Digital Personal Data Protection (DPDP) framework.

“With the DPDP Act making data sovereignty a business imperative, keeping our data and AI workloads within Indian borders lets us scale Claude across the Group with confidence,” said Mohit Kapoor, Chief Technology Officer, Mahindra Group.

CRED’s Head of AI Strategy, Swamy Seetharaman, said local inference would allow the company to combine Claude’s performance with the data governance requirements of the Indian financial sector. “Trust is foundational to every experience we build at CRED, and that increasingly includes the AI infrastructure. Having Claude inference available in India allows us to combine strong performance with the data governance and regulatory expectations of the Indian financial ecosystem,” Seetharaman said.

The National Payments Corporation of India (NPCI) is also using Claude as part of AiNxt, its agentic AI platform. “NPCI is building AiNxt, its agentic AI platform, with Claude - bringing secure, governed AI to enterprise scale,” said Sampath Kumar Rangasamy, Chief Online & New Products Development, NPCI.

Vinayak Godse, CEO of the Data Security Council of India, said Claude Opus-powered agents are being used to probe codebases, systems and infrastructure for vulnerabilities, while reviewing agent workflows, logic and security.

Other enterprises are using Claude across engineering and production workflows. Freshworks said it runs Claude Code across its engineering organisation to accelerate routine build and review work, while Swiggy said Claude was influencing how its engineering teams approach complex technical problems.

“At Swiggy, we’re focused on applying AI to solve complex engineering problems, not just automate routine tasks. Claude is one of the few AI tools we’ve seen influence how our engineering teams approach problem-solving, helping them move faster while maintaining the quality that’s essential at scale,” said Madhusudhan Rao, CTO, Swiggy Ltd.

Invideo, which serves creators in 190 countries, said consistency in output quality across languages and formats was critical to its operations. “Claude gave us consistency in output quality across languages and formats, and Anthropic's team has treated our constraints as their problem rather than ours,” said Sanket Shah, Co-founder and CEO, Invideo.