Anthropic has launched Claude Sonnet 4.5, its latest large language model, and released several upgrades across its AI product lineup.

Claude Sonnet 4.5 now leads benchmarks, such as SWE-bench Verified for software coding and OSWorld for computer use, with a reported 61.4% score.

The company has added features to Claude Code, including checkpoints to save progress, a refreshed terminal interface, and a native Visual Studio Code extension.

Claude’s API now supports longer context editing, and a memory tool for handling complex agent tasks.

The Claude apps can now execute code and create files, like spreadsheets and slides, within conversations.

🔗 Source: Anthropic

Recent Claude developments