Anthropic Launches Claude Sonnet 5: Flagship-Level Performance at a More Affordable Price

Anthropic today officially upgraded its mid-range AI model, launching Claude Sonnet 5 to replace the Sonnet 4.6 released this February, marking the series’ new “best mid-range model.” Within Anthropic’s product lineup, Sonnet sits between the smaller Haiku and the larger Opus, and Sonnet 5 is described by the company as the “most agent-capable Sonnet model yet.”

On performance, Sonnet 5 significantly narrows the gap with the flagship Opus 4.8, achieving comparable results across multiple benchmarks while carrying a more accessible price tag. Starting today, Sonnet 5 will become the default model for both Claude Free and Pro users. Anthropic has also introduced a limited-time promotional rate: $2 per million input tokens and $10 per million output tokens, valid through August 31, 2026.

This update further strengthens Claude’s competitiveness in the mid-range market. By lowering the price barrier and enhancing agentic capabilities, Anthropic is aiming to attract more developers and enterprise users to deploy Sonnet 5 in automation tasks, code generation, and complex reasoning scenarios. As the industry race for model upgrades intensifies, the launch of Sonnet 5 may put noticeable price-to-performance pressure on similar products.