ProductFeb 17, 2026Introducing Claude Sonnet 4.6Sonnet 4.6 delivers frontier performance across coding, agents, and professional work at scale.

Anthropic has released Claude Sonnet 4.6, upgrading its mid-tier model to offer performance levels that rival its previous top-end offerings. The company announced on February 17, 2026, that the new model brings substantial improvements in coding, agent planning, and computer use. By making performance previously exclusive to its more expensive Opus line available at the existing Sonnet price point, this release effectively lowers the cost for businesses to deploy more sophisticated AI for complex, real-world office tasks.

Sonnet 4.6 maintains the same pricing as its predecessor, Sonnet 4.5, starting at $3 per million input tokens, while introducing a 1-million-token context window in beta. Internal testing and early customer feedback indicate a significant leap in capability. For coding tasks, developers preferred Sonnet 4.6 over the November 2025 frontier model, Opus 4.5, in 59% of cases. The model also matches the current-generation Opus 4.6 on the OfficeQA document comprehension benchmark and demonstrates steady gains in the OSWorld benchmark, which measures an AI's ability to operate software like a human using a virtual mouse and keyboard.

The launch of Sonnet 4.6 signals a continued compression in the AI market, where the capabilities of frontier models are rapidly cascading down to more cost-effective tiers. This trend accelerates the economic viability of deploying complex, agentic AI systems at scale for functions like financial analysis, contract routing, and automated software use. As the performance gap between model tiers narrows, it places pressure on competitors and enables a wider range of organizations to move AI applications from experimental phases into production workflows without incurring premium costs.

Anthropic's strategy with Sonnet 4.6 is clear: commoditize yesterday's frontier performance to drive enterprise adoption. By delivering Opus-level capabilities for complex business reasoning and agentic tasks at a Sonnet price, the company is directly targeting the operational cost barrier that has limited the widespread deployment of advanced AI agents in corporate environments.