Sign in
← AI PulseSep 18, 2026

Wire · news · Multi-source brief

Kimi K3 from Moonshot AI Now Available on Amazon Bedrock

Moonshot AI's Kimi K3 model is now available on Amazon Bedrock, offering native vision, a 1-million-token context window, and explicit prompt caching.

By Illumora Editorial

Source · Sep 18, 2026, 4:52 PM · On Illumora · Sep 18, 2026, 4:57 PM

Media from the primary source — shown here so you can stay on Illumora.

Synthesized from multiple allowlisted primaries on the same event. Lanes →

Brief drafted by Illumora’s editorial model from the linked primary source. Ops desk reviews flagged pieces. How we write →

Read the source →AWS Machine Learning Blog — Introducing Kimi K3 on Amazon Bedrock | Amazon Web Services
Save

The Kimi K3 model from Moonshot AI is now available on Amazon Bedrock, providing a new open-weight option for coding and knowledge work. This model includes native vision capabilities and a substantial context window.

Separately, Amazon SageMaker HyperPod Inference Gateway has been introduced as a Kubernetes-native, GPU-aware routing add-on for Amazon EKS.

Key Points

  • Kimi K3 from Moonshot AI is available on Amazon Bedrock.
  • Kimi K3 is an open-weight model.
  • The model offers native vision capabilities.
  • Kimi K3 features a 1-million-token context window.
  • Explicit prompt caching is included to reduce latency and input costs.
  • Amazon SageMaker HyperPod Inference Gateway is a Kubernetes-native, GPU-aware routing add-on for Amazon EKS.
  • This gateway uses real-time GPU signals to route inference requests.
  • It can reduce first-token latency by up to 82% without changes to model servers or client applications.

Context

According to the AWS Machine Learning Blog, Kimi K3 from Moonshot AI provides a new option for coding and knowledge work on Amazon Bedrock. The blog also states that Amazon SageMaker HyperPod Inference Gateway is designed to optimize inference request routing for Amazon EKS.

Why It Matters

The availability of Kimi K3 on Amazon Bedrock offers builders a new open-weight model with advanced features like native vision and a large context window, potentially impacting the efficiency and cost of AI applications. The SageMaker HyperPod Inference Gateway provides a mechanism to reduce inference latency, which can improve the responsiveness of deployed models.

What To Do

  • Review the Kimi K3 documentation on Amazon Bedrock for specific use cases.
  • Compare the pricing and performance of Kimi K3 against other models available on Amazon Bedrock.
  • Investigate the Amazon SageMaker HyperPod Inference Gateway for Amazon EKS deployments to assess potential latency reductions.
  • Note the 82% first-token latency reduction claim for SageMaker HyperPod Inference Gateway in performance evaluations.