New updates on AWS GovCloud (US) | Issue #3

GenAI momentum in AWS GovCloud (US) continues to accelerate, with multiple milestones landing in a single month. Kiro has achieved FedRAMP High and DoD CC SRG IL-4/5 authorization in AWS GovCloud (US) Regions as an agentic engineering platform for secure development workflows. In addition, the following models on Amazon Bedrock have also received FedRAMP High and DoD CC SRG IL-4/5 approval for building and scaling generative AI applications: OpenAI GPT-5.4, OpenAI GPT-OSS (20B and 120B), and NVIDIA Nemotron (Nano 9B v2, Nano 12B v2, Nano 30B, and Super 120B). Alongside these authorizations, Anthropic Claude Opus 4.8, is now available in AWS GovCloud (US) through Amazon Bedrock.

For leaders evaluating AI strategies, these launches enable development teams to use Kiro as their authorized agentic engineering partner for sensitive workloads. Teams also gain access to a broader set of authorized foundation models for deploying generative AI workloads at scale. The OpenAI GPT, OpenAI GPT OSS, and NVIDIA Nemotron models are powered by Mantle, a next-generation distributed inference engine on Amazon Bedrock that provides high-performance serverless inference with zero operator access, automated capacity management, and out-of-the-box compatibility with OpenAI API specifications. In addition to these newly authorized models, Anthropic Claude Opus 4.8 delivers meaningful advances across agentic coding, professional knowledge work, and long-running autonomous tasks, with deeper reasoning, longer autonomous runs, and the consistency to be trusted with production work.

From FedRAMP and DOD SRG IL5 authorizations that remove compliance barriers, to new model families that expand choice, to frontier capabilities that push what's possible in regulated environments, AWS GovCloud (US) is delivering GenAI at a pace that matches the urgency of government modernization. As agencies move beyond experimentation into production-scale AI deployment, the combination of authorized infrastructure, multi-model access, and frontier model performance positions customers to build with confidence rather than compromise. To learn more, visit the Kiro FedRAMP authorization announcement, the AWS GovCloud (US) compliance documentation, or view services in scope for FedRAMP, DoD CC SRG, and FedRAMP Amazon Bedrock Models.

Two new foundation models now join Anthropic Sonnet 4.5 in the Kiro IDE and CLI in the AWS GovCloud (US-West) Region, giving development teams the flexibility to select the model that best fits their workload requirements rather than relying on a single provider. AWS GovCloud (US) is the first AWS Region where OpenAI GPT-5.4 and NVIDIA Nemotron 3 Super 120B are available in Kiro. OpenAI GPT-5.4 brings complex reasoning, coding, document analysis, and multi-step agentic workflows to Kiro, helping developers build AI applications that can interpret context, interact with tools, operate software environments, and verify outputs across multiple steps. GPT-5.4 runs on Amazon Bedrock's next-generation inference engine with isolated queues and durable execution for resilient workloads, available with a 272K context window and a 1.2x credit multiplier.

 NVIDIA Nemotron 3 Super 120B is available as an open weight model option, a hybrid mixture-of-experts architecture activating only 12B of its 120B parameters for high compute efficiency and fast inference on agentic tasks, with a 256K context window and 32K max output at a 0.25x credit multiplier. Teams working on tasks that require deep reasoning across large codebases or multi-step verification may gravitate toward GPT-5.4, while teams prioritizing fast iteration cycles on agentic workloads with cost efficiency in mind may find Nemotron a natural fit. Ensure your IDE or CLI is updated to the latest version, then restart to access the new models from the model selector with no changes to existing workflows or access controls required.

 Learn more: GovCloud Documentation | Kiro Product Page | What's New Post