Anthropic has unveiled its upgraded AI model, Claude 3.5 Sonnet, which brings significant advancements in coding, reasoning, and data analysis, all while upholding the model's dedication to safety and reliability. This new iteration showcases a range of improvements that enhance its functionality and user experience.

One of the standout features of Claude 3.5 Sonnet is its enhanced coding abilities. The model achieves a remarkable 49.0% performance on the SWE-bench Verified, outperforming its predecessors, including Claude 3 Opus. It demonstrates substantial progress in agentic coding and tool use tasks, marking a significant leap forward in AI coding capabilities.

In addition to coding, Claude 3.5 Sonnet introduces new computer use capabilities, currently in beta. This feature allows the AI to navigate computer interfaces much like a human user, with the ability to move cursors, click buttons, and type text. It is the first AI model to offer such computer use in a public beta, scoring 14.9% on the OSWorld evaluation for the screenshot-only category, which is notably higher than the next-best AI system at 7.8%.

Performance enhancements are another key aspect of Claude 3.5 Sonnet. The model operates at twice the speed of Claude 3 Opus for most workloads, providing improved instruction following and natural language understanding. Users can expect enhanced reliability and consistency in the AI's responses, making it a more dependable tool for various applications.

Despite its advanced capabilities, the model maintains Anthropic's commitment to safety, remaining at AI Safety Level 2 (ASL-2), which indicates that appropriate safeguards are in place for its current deployment. Anthropic has conducted extensive safety evaluations, collaborating with the US and UK AI Safety Institutes to ensure responsible development.

Claude 3.5 Sonnet is designed for real-world applications, excelling in complex tasks such as context-sensitive customer support and orchestrating multi-step workflows. Its improved capabilities in coding, computer use, and general reasoning are expected to drive new innovations across various industries.

The release of Claude 3.5 Sonnet underscores Anthropic's ongoing commitment to developing more capable AI systems while maintaining high standards for safety, reliability, and cost-effectiveness. This latest model is poised to make a significant impact, offering users a powerful tool for a wide range of applications.

Try Now