Anthropic Releases Claude Sonnet 5 with Enhanced Agentic Capabilities
Original: Claude Sonnet 5
Why This Matters
Sonnet 5 democratizes agentic AI by delivering Opus-class performance at mid-tier pricing, expanding autonomous AI access for developers.
Anthropic announced Claude Sonnet 5 on June 30, 2026, featuring advanced agentic capabilities including planning, tool use, and autonomous operation. Performance approaches Opus 4.8 at lower cost, with introductory pricing of $2/$10 per million tokens through August 31.
Anthropic introduced Claude Sonnet 5, positioned as the most agentic Sonnet model to date. The model can plan, use tools like browsers and terminals, and operate autonomously at capability levels that previously required larger, more expensive models. Sonnet 5 demonstrates substantial improvements over its predecessor Sonnet 4.6 in reasoning, tool use, coding, and knowledge work. Performance metrics show Sonnet 5 approaching Opus 4.8's capabilities while maintaining cost advantages. Safety assessments indicate Sonnet 5 exhibits lower rates of undesirable behaviors than Sonnet 4.6 and shows reduced ability to perform cybersecurity tasks compared to Opus models. The model is available across all Claude plans: default for Free and Pro users, with access for Max, Team, and Enterprise customers. Pricing begins at $2 per million input tokens and $10 per million output tokens through August 31, 2026, rising to $3 and $15 respectively thereafter. Claude Sonnet 5 is accessible via Claude API, Claude Code, and the Claude Platform. Performance evaluations on agentic search (BrowseComp) and computer use (OSWorld-Verified) show Sonnet 5 provides wider cost-performance options than Sonnet 4.6, with higher-effort performance sometimes matching Opus 4.8. Early access partners reported Sonnet 5 successfully completes complex tasks where previous versions fell short.