Opus 4.8 is now available in Kiro
Nima Kaviani
Product
Starting today, Claude Opus 4.8 is available in Kiro IDE, CLI, and Web. It's Anthropic's most intelligent Opus model, a direct upgrade from Opus 4.7, with stronger self-verification, more efficient tool calling, and better follow-through on long-horizon projects.
Across the IDE, CLI, and Web, Opus 4.8 delivers sharper judgment, tighter tool orchestration, and more reliable follow-through on the kind of complex work that used to require close supervision. It's a meaningful step up in what you can confidently delegate.
On agentic coding benchmarks, Opus 4.8 scores 69.2% on SWE-Bench Pro, up from 64.3% for Opus 4.7. But beyond the numbers, the behavior change is what you'll actually feel. Opus 4.8 plans before it edits, catches its own mistakes, and finds creative paths around obstacles instead of stalling. It pushes back when a plan isn't sound and flags uncertainties rather than making unsupported claims. It feels less like a code-completion engine and more like a senior engineer - one that holds context, checks its work, and follows through across long sessions.
For spec-driven workflows in the IDE, that means higher-fidelity implementation across broader changes. For the CLI and Web, it means complex multi-step tasks run to completion more reliably, less drift, fewer dead ends, and less need to supervise closely.
Opus 4.8 is available with experimental support to Kiro Pro, Pro+, and Power customers logging in with AWS IAM Identity Center, Google, GitHub, and AWS Builder ID. It ships with the full 1M context window and a 128K max output token limit.
Opus 4.8 is available in the AWS US-East-1 (Northern Virginia) and AWS Europe (Frankfurt) regions, with cross-region inference support. It carries a 2.2x credit multiplier, the same as Opus 4.7.
Download Kiro or restart the IDE or CLI to check for the latest available models. For the Kiro CLI users to be able to get the best experience out of Opus 4.8, they need to be on Kiro CLI v2.5.0 or above.