Claude Sonnet 4.5 is the newest AI model from Anthropic, and it’s built for people who want to create smarter apps and digital assistants. It belongs to the Claude 4 family, and you can use it through Anthropic’s website, their mobile app, or even a command-line tool called Claude Code if you like working in the terminal.
So what’s different about this version? Claude Sonnet 4.5 is better at coding than its predecessor. On a test called SWE-bench Verified, which checks how well an AI can actually solve real software problems, it scored 77.2%. That’s a jump from the previous model’s 72.7%. It’s also much better at handling everyday computer tasks, like browsing websites or working with spreadsheets, scoring 61.4% on the OSWorld benchmark (up from 42.2%). Anthropic says this model can keep working on tough problems for more than 30 hours at a stretch. They’ve also added something called the Claude Agent SDK, which helps developers manage memory and coordinate between different parts of an AI system. The price hasn’t changed: $3 for a million input tokens, and $15 for a million output tokens.
Why does any of this matter? If you’re someone who wants to build your own AI agents, or you’re working on tools that need to run for a long time without stopping, these upgrades will save you a lot of trouble. The new features help solve problems that most teams would otherwise have to figure out on their own. But if you’re just looking for a chatbot to talk to, you probably won’t notice much difference. This update is really about making AI that can work on its own for hours, not just answer questions in a chat window.
