SMART WORLDJOURNALSubmit Manuscript
← All technology news

Artificial Intelligence

Moonshot AI's Kimi Code CLI v2.7.0 Integrates Native AWS Credential Support for Amazon Bedrock Models

Moonshot AI updates Kimi Code CLI to version 2.7.0, enabling seamless authentication with Amazon Bedrock models through native AWS credential chains. This eliminates the need for separate API keys, streamlining secure access to foundation models like Kimi K3.1 Pro Highspeed via AWS standard workflows.

NEWS

Moonshot AI has released version 2.7.0 of its open source Kimi Code CLI, introducing a native Amazon Bedrock provider that integrates directly with AWS’s standard credential chain. Previously, users of Kimi Code interfacing with Amazon Bedrock models were required to manually manage separate Bedrock API keys. This update simplifies authentication by allowing use of standard AWS credentials, including named profiles and single sign-on (SSO), enhancing security and administrative convenience.

Amazon Bedrock is a service providing pay-per-token inference access to a variety of open and proprietary AI models, including Moonshot AI’s own Kimi series. Models like Kimi K3.1 Pro Highspeed support advanced features such as streaming responses, configurable reasoning, image input handling, and client-side tool calling. Bedrock processes requests within AWS’s data boundary, adhering to strong governance and compliance standards.

The new native provider in Kimi Code CLI works by following AWS’s credential resolution chain to obtain temporary credentials. It uses the @aws/bedrock-token-generator library to create ephemeral Bedrock API keys valid for up to 12 hours or the duration of the AWS session. Requests are routed to the appropriate regional Bedrock endpoints inferred from model IDs, supporting protocol overrides where needed, such as when interfacing with VPC or Mantle endpoints.

This approach consolidates authentication management, eliminating the need to manually create, distribute, or rotate separate Bedrock API keys. It leverages existing AWS identity access management (IAM) policies, which can be scoped with the principle of least privilege to maintain security. The CLI supports specifying profiles and regions explicitly, optimizing for various deployment scenarios including global cross-region inference with cost benefits.

Users must configure the Kimi Code CLI with the new provider and declare model capabilities explicitly, as automatic detection is limited. For example, capabilities such as 'always_thinking', 'thinking', 'image_in', and 'tool_use' must be enabled based on the model card. While the native provider streamlines authentication, it remains essential to properly configure AWS credentials and validate IAM permissions for seamless operation.

The update not only enhances security by relying on short-lived AWS credentials recommended by AWS best practices but also improves usability for developers working with Bedrock models through Kimi Code CLI. However, it requires users to maintain updated CLI versions, manage AWS profiles carefully, and understand configuration nuances such as region settings and endpoint overrides.

Moonshot AI’s native AWS credential support within Kimi Code CLI marks a significant improvement in developer experience for AI model inference on Amazon Bedrock. It reduces credential sprawl and administrative overhead while maintaining compliance and security standards, facilitating broader adoption of foundation models in cloud-based applications.

Source & further reading

Use Kimi Code CLI with models on Amazon Bedrock ↗

Source published 2026-11-20. Claims and limitations should be checked against the original announcement or research.