Anthropic releases Claude Sonnet 5 with better reasoning, coding, and agentic performance

Anthropic has introduced Claude Sonnet 5, a major upgrade to its mid-tier AI model line with stronger agentic capabilities. The model can plan tasks, use external tools, work with browsers and terminals, and operate with more autonomy. Anthropic describes it as its most agentic Sonnet model so far, bringing capabilities that were previously more associated with larger systems. Claude Sonnet 5 approaches the performance of Anthropic’s flagship Claude Opus 4.8 model while costing less to run. It improves on Claude Sonnet 4.6 across reasoning, coding, and knowledge-work benchmarks. API pricing starts at a promotional rate of $2 per million input tokens and $10 per million output tokens until August 31, after which standard pricing will apply. Anthropic also reports fewer undesirable behaviors, reduced hallucinations, and stronger resistance to prompt-injection attacks compared with the previous generation, making Sonnet 5 safer for agentic use cases. The company noted the model's cybersec...