Anthropic released Claude Sonnet 4.5, calling it the best coding model in the world and the strongest for building complex agents, and highlighting its ability to hold focus on multi-step work for more than 30 hours. It scored 77.2% on SWE-bench Verified, averaged over 10 trials with a large thinking budget, and Anthropic described it as its most aligned frontier model to date. Shipping alongside were checkpoints in Claude Code for rolling back work, a native VS Code extension, and the Claude Agent SDK exposing the infrastructure behind Claude Code to developers. The Claude for Chrome extension went to Max subscribers, and the Claude apps gained code execution and the ability to create spreadsheets, slides, and documents. Pricing stayed at Sonnet 4 levels of $3 and $15 per million tokens.