Amazon Bedrock Adds India-Based Inference for Kimi K3

What happened

Amazon Bedrock now supports in-country inference for Kimi K3 in its Mumbai and Hyderabad Regions through the in.moonshotai.kimi-k3 profile. AWS explains how to select geographic inference and demonstrates a pipeline for extracting fields from Indic-language documents and answering follow-up questions in Hinglish. The sample uses synthetic documents, and AWS advises verifying extraction outputs before relying on them for important decisions.

Read next

• More changes in today’s briefingOpenAI’s updated Model Spec explained through its five core values

The story so far

AWS now supports in-country inference for Kimi K3 on Amazon Bedrock in the Mumbai and Hyderabad Regions through the in.moonshotai.kimi-k3 profile. AWS also published a sample pipeline that uses optical character recognition to extract fields from Indic-language documents and answer follow-up questions in Hinglish. The sample uses synthetic documents, and AWS advises checking extraction results before relying on them for important decisions. For less latency-sensitive workloads, the flex tier costs 50% less than default standard pricing; the priority tier offers higher queue priority at a higher price for latency-critical workloads.