Anthropic Launches Claude Sonnet 5, Advanced AI Model for Autonomous Tasks

Anthropic has introduced Claude Sonnet 5, a new artificial intelligence model designed to handle complex autonomous tasks with capabilities previously reserved for larger, more expensive systems. The model can plan workflows, operate tools such as browsers and terminals, and perform sophisticated work independently.

Claude Sonnet 5 represents a significant advancement in the Sonnet model line. While earlier versions like Sonnet 3.5, 3.6, and 3.7 first demonstrated strong capabilities in coding and tool usage, recent performance gains have been concentrated in Opus-class models. Sonnet 5 closes this gap by delivering performance comparable to Opus 4.8 at substantially lower costs.

Performance improvements include marked enhancements across reasoning, tool use, coding, and knowledge work tasks compared to its predecessor, Sonnet 4.6. Testing shows Sonnet 5 exhibits fewer undesirable behaviors than earlier versions and maintains improved safety profiles for autonomous applications. The model demonstrates reduced capability for cybersecurity tasks compared to current Opus offerings.

The new model is available across all subscription tiers immediately. Free and Pro users receive it as their default model, while Max, Team, and Enterprise customers can access it as an option. Through August 31, 2026, introductory pricing is set at $2 per million input tokens and $10 per million output tokens. Standard pricing after that date will be $3 per million input tokens and $15 per million output tokens. Developers can access Claude Sonnet 5 through the Claude API using the identifier claude-sonnet-5.

Early access feedback from partner organizations highlights consistent strengths. Users report the model completes complex tasks where previous Sonnet versions would stop prematurely, performs self-verification without explicit instruction, and delivers these capabilities at attractive pricing. Partners have noted particular effectiveness in software engineering workflows, automation tasks, legal research, data analysis, and insurance operations.

Performance evaluations indicate Sonnet 5 provides superior cost-efficiency at medium effort levels, with higher-effort performance potentially matching Opus 4.8 on certain tasks, offering users flexibility in balancing cost against capability requirements.