Coding, planning, reasoning: Claude Sonnet 4.6 does it all

In a statement, the company described it as a comprehensive upgrade across coding, computer use, long-context reasoning, agent planning, knowledge work and design. The model also introduces a 1 million-token context window in beta, significantly expanding its ability to process and analyse large volumes of information.

For users on the Free and Pro plans, Claude Sonnet 4.6 is now the default model on Claude.ai and Claude Cowork. Pricing remains unchanged from Sonnet 4.5, starting at $3 per million input tokens and $15 per million output tokens.

Anthropic said Sonnet 4.6 delivers substantial improvements in coding performance, along with enhanced consistency and instruction-following capabilities. Developers with early access have reportedly shown a strong preference for the new model over its predecessor — and, in many cases, even over the company’s flagship model from November 2025, Claude Opus 4.5.

The company noted that tasks previously requiring an Opus-class model — including real-world, economically valuable office workflows — can now be handled effectively by Sonnet 4.6. The model also demonstrates marked improvements in computer-use capabilities compared to earlier Sonnet versions.

On safety, Anthropic said it conducted extensive evaluations prior to release. Researchers concluded that Sonnet 4.6 is as safe as, or safer than, recent Claude models. According to the company, safety assessments found the model to have “a broadly warm, honest, prosocial, and at times funny character, very strong safety behaviors, and no signs of major concerns around high-stakes forms of misalignment.”