I tried Claude Opus 5.5 which became selectable in Kiro
This page has been translated by machine translation. View original
Introduction
On September 28, 2026, Claude Opus 5.5 became available as an Experimental preview in Kiro for enterprise licenses.
This time, I was also able to confirm its operation in Kiro IDE / CLI, so I'll introduce it here.
Overview of Claude Opus 5.5
According to Anthropic's announcement, Opus 5.5 delivers performance comparable to Fable 5.1 on many tasks.
The API pricing is $4 for Input / $20 for Output (per 1 million tokens), which is 20% cheaper than Opus 5.
Test Environment
Verified with Kiro CLI 2.24.1 (Linux aarch64).
Model Selection in Kiro IDE
In Kiro IDE, clicking the model selector at the bottom of the screen (shown as "Auto" in the image, at the arrow position) opens the model selector. Claude Opus 5.5 appears in the list as an Experimental preview. The context window is 1M, and the credit multiplier is 2x. Both Opus 5 and Opus 4.8 are 2.2x.
Model Selection in Kiro CLI
In Kiro CLI, use /model to switch models. claude-opus-5.5, like in the IDE, is 2.00x credits. This is a lower multiplier than the 2.20x for claude-opus-5 and claude-opus-4.8 listed alongside it.
› /model
Select model: type to search
❯ auto 1.00x credits Models chosen by task for optimal usage and consistent quality
claude-opus-5.5 2.00x credits Experimental preview of Claude Opus 5.5 model with 1M context window [active]
claude-opus-5 2.20x credits Claude Opus 5 model with 1M context window
claude-sonnet-5 1.30x credits Claude Sonnet 5 model with 1M context window
claude-opus-4.8 2.20x credits Claude Opus 4.8 model with 1M context window
gpt-5.6-sol 4.40x credits Experimental preview of OpenAI GPT 5.6 Sol with 1M context window
gpt-5.6-terra 2.20x credits Experimental preview of OpenAI GPT 5.6 Terra with 1M context window
gpt-5.6-luna 1.10x credits Experimental preview of OpenAI GPT 5.6 Luna with 1M context window
The reduction in multiplier is approximately 9%. This is a smaller change compared to the 20% decrease in Anthropic API Input/Output unit pricing relative to Opus 5.
Model Governance Selection in Kiro
For this account, model governance had been enabled in order to use Fable 5.1. In this state, after checking Opus 5.5 in the admin screen, Opus 5.5 was added to the model selection in both CLI and IDE. If Opus 5.5 doesn't appear as an option, please also check the model governance settings.
From "Model Availability" in the admin screen settings, open "Manage Approved List."
Check the box for Claude Opus 5.5 in the list of available models.
Summary
Claude Opus 5.5 is now available for selection in Kiro as well.
The following comment has been shared by Kiro regarding the announcement of Opus 5.5.
"Every call an agent makes is time and cost a developer feels. On a public benchmark of real command-line tasks, Claude Opus 5.5 solved more than Opus 5 while making about 40% fewer calls and using half the tokens. For developers building with Kiro, that means faster, more affordable agent sessions for routine tasks and complex challenges alike. Opus 5.5 will soon be available in Kiro."
— Deepak Singh, VP of Agentic AI, Kiro（Introducing Claude Opus 5.5）
Regarding the performance improvement of approximately 40% fewer calls and half the token usage, we plan to evaluate this going forward.
If Opus 5.5 appears in your list of available models in Kiro, please give it a try.
