Introducing India-based inference for Kimi K3 on Amazon Bedrock

Amazon Bedrock now supports in-country inference in India for Moonshot AI's Kimi K3, letting customers run workloads that have strict data residency requirements.

• Kimi K3 is a multi-modal, open-weight foundation model that provider Moonshot AI reported achieved frontier-level coding and general agentic intelligence at release.
• The model is offered through cross-Region inference profiles, so a request made from a source Region is processed in one of several eligible destination Regions, with data encrypted in transit and kept on the AWS network.
• The inference profile ID controls which destination Regions are eligible, and the India profile in.moonshotai.kimi-k3 keeps inference inside Indian Regions.
• Ultrahuman's VP of Engineering, Aditya Patawari, says the company's Jade health assistant handles more than 14,000 conversations a day on Kimi K3, with health data staying inside AWS and Mumbai as a fallback Region.
• Patawari says prompt caching cut LLM costs by 80 percent and made multi-turn chats 85 percent faster.
• The post walks through an Indic-language document use case, reading scanned Hindi documents and answering follow-up questions in code-mixed Hinglish, with source code on GitHub.