Claude Code just got more convenient with a new fallback model setting!
New Features and Fixes in Claude Code v2.1.166
Original: v2.1.166
Importance: 新機能とバグ修正が含まれ、多くのユーザーに影響するため。
Summary
Anthropic's Claude Code has been updated to v2.1.166, introducing new features and bug fixes. A new `fallbackModel` setting allows up to three fallback models to be tried in order when the primary model is overloaded or unavailable. Other updates include strengthened cross-session messaging, improved behavior on API errors, and fixes for terminal display issues, enhancing user experience with Claude Code.
Key Points
- Addition of fallback model setting
- Strengthened cross-session messaging
- Improved behavior on API errors
- Fixes for terminal display issues
- Bug fixes enhance usability
View developer notes (APIs, breaking changes, migration)
In Claude Code v2.1.166, the new `fallbackModel` setting allows up to three fallback models to be tried in order when the primary model is overloaded or unavailable. Additionally, settings like `MAX_THINKING_TOKENS=0` and `--thinking disabled` disable thinking on models that think by default. The `claude update` command now announces the target version before downloading, and a feature to retry once on the fallback model for API errors is added. Various bugs and display issues in terminals have also been fixed, enhancing usability.
Source: https://github.com/anthropics/claude-code/releases/tag/v2.1.166
Outlet: Claude Code Releases
This article is an AI-generated summary (OpenAI GPT-4o-mini) of publicly available information from Anthropic, OpenAI, Google, Meta, Mistral, DeepSeek, Sakana, and other vendors. The original source URL is always provided in accordance with fair-use citation requirements. Summaries are AI-generated and may contain mistranslations or misinterpretations. Always verify details with the original source.