Anthropic Releases Claude Sonnet 5 with 60% Lower Cost Than Opus on June 30

On June 30, Anthropic released Claude Sonnet 5, a mid-tier model delivering performance nearly equivalent to its flagship Opus 4.8 at approximately 60% lower cost. The model achieved 63.2% on the SWE-bench Pro coding agent benchmark, surpassing Sonnet 4.6's 58.1% and approaching Opus 4.8's 69.2%. On the GDPval-AA v2 knowledge work benchmark, Sonnet 5 scored 1618 points, exceeding Opus 4.8's 1615.

Initial API pricing is set at $2 per million input tokens and $10 per million output tokens through August 31; after the promotional period, rates will increase to $3 and $15 respectively, remaining significantly below Opus 4.8's $5 and $25.