Amazon Bedrock Integrates Moonshot AI's 2.8-Trillion Parameter Kimi K3 Model
Amazon Bedrock has officially integrated Moonshot AI's Kimi K3 model, a 2.8-trillion parameter open-weights model featuring native vision capabilities and a 1-million-token context window. It also becomes the first open-weights model on Amazon Bedrock to offer explicit prompt caching functionality. This integration gives AWS enterprise users access to a massive open-weights model tailored for long-context coding and multi-document analysis while maintaining enterprise security and compliance standards. The addition of explicit prompt caching helps significantly lower latency and token input costs for repetitive queries. Kimi K3 operates within Amazon Bedrock's security boundaries, utilizing unified access control, data encryption, and audit mechanisms. By caching reusable prompt contexts across multiple API calls, developers can optimize inference performance for complex workflows.
## BACKGROUND
Amazon Bedrock is AWS's fully managed service that provides access to leading foundation models through a single API with enterprise security features. Prompt caching is an optimization technique that stores pre-processed context—such as repetitive system instructions or large reference documents—so the model does not need to reprocess those tokens on subsequent calls.