Anthropic Presents Advanced Programming Models at the Developers Conference

Anthropic has introduced two new models, Claude Opus 4 and Claude Sonnet 4, at a recent conference for developers. Claude Opus 4 demonstrated exceptional performance on the SWE-bench validation set, achieving a top score of 72.5% and reaching 79.4% in high-computation mode, positioning itself as a leading automatic programming model worldwide. Claude Sonnet 4 also performed impressively with a score of 72.7%, surpassing OpenAI's o3 and Codex-1 models. Tests conducted by Rakuten revealed that Opus 4 can program continuously for seven hours while efficiently handling complex tasks, setting a new standard in the industry. The new models support the use of tools in parallel and feature enhanced memory mechanisms, with Claude Code now fully accessible.