Anthropic Sets Claude Sonnet 5 as Default in Claude Code, Stirring Debate Over Capability Trade-offs
Key Takeaways
- ▸Claude Sonnet 5 is now the default model in Claude Code v2.1.197, featuring a native 1M-token context window
- ▸Promotional pricing of $2/$10 per Mtok is available through August 31st, offering cost savings compared to standard rates
- ▸The change prioritizes accessibility and efficiency over raw capability, potentially favoring speed and cost over complex reasoning performance
Summary
Anthropic has made Claude Sonnet 5 the default model in Claude Code (v2.1.197), replacing what users suggest was a more capable default option. The new Sonnet 5 model includes a native 1M-token context window and arrives with promotional pricing of $2/$10 per Mtok through August 31st. While the update offers expanded context capacity and cost efficiency, the change has prompted discussion among users who view Sonnet as a less capable alternative for complex reasoning tasks, despite the pricing advantages. The shift reflects a broader trade-off between performance and accessibility in the AI tooling market.
The release notes emphasize the expanded context window and promotional pricing as key benefits for Claude Code users. Additional improvements in v2.1.197 address organization-level model defaults, UI enhancements for file attachments, and various security and stability fixes. However, the decision to make Sonnet the default—rather than the more powerful Opus model—appears to have raised questions about whether price optimization is coming at the expense of model quality for the typical use case.
- Users can still select alternative models, including the more capable Claude Opus, via the /model command or organization settings
