Anthropic Claude 2

Claude 2 is the newest and most capable version yet. Claude 2 offers stronger performance, more thoughtful and detailed responses, and is now available both through an API and a new public beta site at claude.ai. Users have shared that Claude is easy to talk to, provides clear explanations, handles longer conversations well, and avoids harmful or inaccurate output. This release includes improvements in coding assistance, math ability, and reasoning skills. For instance, the newest version scored 76.5% on the multiple-choice section of the Bar exam (up from 73.0% with Claude 1.3). Claude 2 also performs at a very high level on standardized tests, landing above the 90th percentile on GRE reading and writing and around the middle of the pack on quantitative reasoning.

Claude is designed to feel like a helpful teammate — someone you can speak to naturally, who can assist with a wide range of tasks. The new version is available to businesses at the same price as Claude 1.3, and people in the US and UK can try the beta chat experience today.

In this version, You can now submit very large inputs — up to roughly 100K tokens — which means Claude can review lengthy documents, large technical files, or even whole books at once. It can also produce longer, polished writing such as letters, memos, and creative stories in a single pass, making it more useful for complex projects and long-form content.