Anthropic has introduced Claude Sonnet 5, its latest AI model designed to deliver advanced agentic capabilities, enabling developers to build AI applications that can independently plan tasks, use external tools, and execute complex workflows with greater efficiency.

According to the company, Claude Sonnet 5 significantly improves reasoning, coding, tool use, and knowledge work compared to its predecessor, Claude Sonnet 4.6. Anthropic says the model now delivers performance approaching its more powerful Opus 4.8 model while offering substantially lower operating costs, making advanced agentic AI more accessible to enterprises and developers.

The new model is designed to autonomously perform multi-step tasks using tools such as web browsers and terminals, reflecting the industry's growing shift toward AI agents capable of executing workflows rather than simply responding to prompts.

Anthropic also highlighted safety improvements. Internal evaluations indicate that Claude Sonnet 5 demonstrates a lower rate of undesirable behaviours than Sonnet 4.6 and is generally safer for deployment in agentic environments. The company noted that the model also has a lower capability for performing cybersecurity-related tasks compared with its Opus models, reducing certain deployment risks.

Claude Sonnet 5 is now available across all Claude subscription tiers, serving as the default model for Free and Pro users while also supporting Max, Team, and Enterprise customers. The model is available through Claude Code and the Claude API.

To encourage developer adoption, Anthropic has introduced introductory API pricing of US$2 per million input tokens and $10 per million output tokens until August 31, 2026. After the promotional period, pricing will increase to $3 per million input tokens and $15 per million output tokens, remaining significantly below the pricing of the company's Opus models.

Early access partners reported that Claude Sonnet 5 completes complex, multi-step tasks more reliably than previous Sonnet models, proactively validates its own outputs, and delivers stronger cost-performance efficiency. The launch underscores the growing competition in the AI industry to deliver highly capable, enterprise-ready AI agents that combine advanced reasoning with affordable deployment economics.