Claude Sonnet 4.6

About Claude Sonnet 4.6

Claude Sonnet 4.6 represents Anthropic's most capable Sonnet model to date, released on February 17, 2026. This full upgrade enhances the model's skills across coding, computer use, long context reasoning, agent planning, knowledge work, and design. With a 1 million token context window available in beta on the API, Sonnet 4.6 can process entire codebases, lengthy contracts, or dozens of research papers in a single request while maintaining effective reasoning across all context.

The model delivers performance previously requiring Opus class models, including on economically valuable office tasks. Early testing showed developers preferred Sonnet 4.6 over Sonnet 4.5 roughly 70% of the time, citing better context reading before code modification and reduced logic duplication. Users even preferred it to Opus 4.5 from November 2025 approximately 59% of the time, noting significant reductions in overengineering and "laziness" while improving instruction following and multi step task consistency.

Key Features

• 1M Token Context Window: Process entire codebases, lengthy contracts, or research papers in a single request with effective long context reasoning across all provided material.
• Advanced Computer Use: Navigate digital environments with 72.5% accuracy on OSWorld Verified benchmark, enabling automation of browser-based workflows and multi step form completion.
• Frontier Coding Performance: Achieve 79.6% on SWE-bench Verified and 59.1% on Terminal Bench 2.0, with superior instruction following and reduced code duplication.
• Hybrid Reasoning: Produce near instant responses or extended step by step thinking with API control over thinking effort levels.
• Agentic Capabilities: Superior tool selection, error correction, and sustained coherence for autonomous AI workflows and customer facing agents.
• Enterprise Integration: Available via Claude API, Amazon Bedrock, Google Cloud Vertex AI, and Microsoft Foundry with prompt caching and batch processing support.