Customers can now use Claude Sonnet 4.5 in Amazon Bedrock in AWS GovCloud (US-West) and AWS GovCloud (US-East) via US-GOV Cross-Region Inference. Claude Sonnet 4.5 is Anthropic’s most intelligent model, excelling at building complex agents, coding, and long-horizon tasks while maintaining optimal speed and cost-efficiency for high-volume use-cases.
Claude Sonnet 4.5 currently leads the SWE-bench Verified benchmarks with enhanced instruction following, better code improvement identification, stronger refactoring judgment, and more effective production-ready code generation. This model excels at powering long-running agents that tackle complex, multi-step tasks requiring peak accuracy—like autonomously managing multi-channel marketing campaigns or orchestrating cross-functional enterprise workflows. In cybersecurity, it can help teams shift from reactive detection to proactive defense by autonomously patching vulnerabilities. For financial services, it can handle everything from analysis to advanced predictive modeling.
Through the Amazon Bedrock API, Claude can now automatically edit context to clear stale information from past tool calls, allowing you to maximize the model’s context. A new memory tool lets Claude store and consult information outside the context window to boost accuracy and performance.
To get started with Claude Sonnet 4.5 in Amazon Bedrock, read the News Blog, visit the AWS GovCloud (US) console console, Anthropic’s Claude in Amazon Bedrock product page, and the Amazon Bedrock pricing page.
Categories: general:products/amazon-bedrock,marketing:marchitecture/artificial-intelligence,general:products/aws-govcloud-us
Source: Amazon Web Services
