Anthropic Unveils Claude Sonnet 5 AI Model with Enhanced Autonomous Capabilities and Lower Costs

Anthropic announced the launch of Claude Sonnet 5, its latest mid-tier artificial intelligence model, which now serves as the default for both free and Pro users on its platform. The company describes this version as its most advanced "Agentic AI" model, capable of operating independently and handling complex tasks without close human supervision.

This release marks a significant shift in the global AI creator market, moving focus from simple chatbots to fully autonomous AI agents. Unlike previous models, Claude Sonnet 5 can perform long-term planning, utilize digital tools such as web browsers and command terminals, and complete multi-step tasks that earlier versions struggled to finish due to limited reasoning abilities.

A key improvement is the model's self-monitoring and error-correction capability, allowing it to automatically review and fix its outputs without user intervention. This feature, previously found only in more expensive and resource-heavy models, enables developers to integrate Claude Sonnet 5 into complex applications while maintaining high accuracy and reliability.

Anthropic had concentrated on enhancing its Opus and Mythos models but faced regulatory and export restrictions from the U.S. government. By strengthening the Sonnet series, the company offers a powerful and accessible alternative that narrows the gap with major competitors OpenAI and Google, while adhering to strict safety and cybersecurity standards by default.

Performance-wise, Claude Sonnet 5 significantly outperforms its predecessor Sonnet 4.6 and approaches the capabilities of the premium Opus 4.8 model. In coding tests, it scored 63.2%, placing it among the leading AI technologies worldwide. Despite its power, the model is priced substantially lower than competitors, at $2 per million input tokens and $10 per million output tokens during a limited launch pricing valid through August.

Anthropic has also increased capacity for its tools like Claude Code and Cowork, expanded chat interface usage limits, and improved response speed and accuracy. End users will notice faster, more reliable answers, making Claude Sonnet 5 a stable and secure tool for everyday work.