On October 9, 2026, AWS announced that Anthropic’s latest models, Claude Opus 5.5 and Claude Sonnet 5.5, are now accessible within the Kiro IDE and CLI for AWS GovCloud (US) Regions. The integration brings improved speed, token efficiency, and massive context windows to government developers and regulated industries using the AI coding environment.
What was announced
AWS has deployed two new Anthropic models to Kiro, its AI-assisted developer toolset. Users operating within the highly secure AWS GovCloud (US) environment can now select either Claude Opus 5.5 or Claude Sonnet 5.5 directly from the Kiro model selector. To access the upgrade, developers must update their IDE or CLI to the latest version and restart the application.
Key details
The update introduces distinct performance profiles for different coding workloads, though both models share a massive 1 million token context window. According to the AWS announcement, the new models offer the following capabilities:
- Claude Opus 5.5: Billed as Anthropic’s most capable Opus model to date, it is specifically designed to handle long-running agentic coding tasks. The model features adaptive thinking, dynamically determining how much effort to apply to each specific request. Based on Kiro’s internal testing, Opus 5.5 completes tasks using approximately 40% fewer tool calls and about half the tokens compared to Opus 5. It carries a 2.0x credit multiplier, which is a reduction from Opus 5’s 2.2x rate.
- Claude Sonnet 5.5: Positioned as the fastest Sonnet model yet, it delivers an output speed more than 30% faster than its predecessor. AWS notes that its performance on real-world coding and knowledge tasks approaches that of Opus 5.5, while also featuring clearer writing. It operates with a lower 1.3x credit multiplier.
Why it matters
For developers restricted to secure, compliance-heavy cloud environments, the addition of the 5.5 generation models offers a powerful combination of increased capability and improved cost efficiency. The massive 1 million token context window allows developers to feed entire repositories, extensive compliance documentation, and complex system logs into the AI at once, drastically improving the assistant’s situational awareness.
The “adaptive thinking” of Opus 5.5 further streamlines workflows by significantly reducing the sheer volume of tokens and tool calls needed to complete complex, multi-step agent-driven tasks. Combined with the lowered credit multiplier, teams can execute intensive coding automation more affordably. Meanwhile, Sonnet 5.5 provides a reliable “cost-performance dial,” allowing developers to prioritize high throughput and faster generation without severely sacrificing reasoning capabilities.
Limitations and open questions
The stated performance metrics—most notably roughly 40% fewer tool calls and about half the tokens for Opus 5.5—are based entirely on Kiro’s internal benchmarking. Furthermore, the announcement details usage costs strictly in terms of Kiro’s internal credit multipliers, leaving exact dollar-value expenditures dependent on a user’s specific AWS GovCloud contract and broader infrastructure usage tiers.
Sources
- Anthropic Claude Sonnet 5.5 and Claude Opus 5.5 are now available on Kiro in AWS GovCloud (US) — AWS What's New (official announcement), October 9, 2026
Related reading
- AI to Empower Developers, Says Embracer Group
- Intel and AMD’s Copilot Plus PCs to Launch Without Copilot AI Features
This article was prepared with AI assistance from the sources listed above and is subject to our editorial policy. Spot an error? Request a correction.