Anthropic’s new Claude Opus 5.5 and Claude Sonnet 5.5 models are now available in the Kiro development environment for AWS GovCloud (US). Opus 5.5 adapts effort per request, using about 40% fewer tool calls and half the tokens of Opus 5 in internal tests. Sonnet 5.5 runs more than 30% faster and brings coding performance close to Opus 5.5 while keeping a lower credit cost. Both models support a 1 million token context window; Opus 5.5 has a 2.0× credit multiplier and Sonnet 5.5 a 1.3× multiplier. Users must update Kiro to the latest version and restart to see the new options.
Why it matters
Government developers can now run higher‑capability coding assistants at lower cost within the secure GovCloud environment.
