Claude Sonnet 4.6 Update.

Anthropic announced the release of Claude Sonnet 4.6, a new model positioned to deliver high-end performance for enterprise coding, agentic workflows, and direct computer automation. The update is significant as it makes capabilities previously exclusive to the company's premium Opus series available at a much lower price point. Sonnet 4.6 is now the default model for users on Anthropic's Free and Pro plans, indicating a strategy to broaden access to its more advanced AI for scaled, practical business applications.

According to the company, Sonnet 4.6 shows substantial improvements in coding, instruction following, and the ability to use software through a simulated mouse and keyboard, as measured by the OSWorld benchmark. Pricing remains the same as its predecessor at $3 per million input tokens and $15 per million output tokens, despite performance that early users reportedly preferred over the previous-generation Opus 4.5 model. The model also features a 1 million token context window in beta and enhanced resistance to prompt injection attacks, a key security concern for AI agents interacting with external websites.

This release intensifies market competition by narrowing the performance gap between mid-tier and flagship AI models, putting pressure on competitors' pricing structures. By pushing frontier-level reasoning and computer-use skills into a more cost-effective package, Anthropic is aiming to accelerate the shift from conversational AI to functional AI agents that can perform complex, multi-step tasks within enterprise software. The focus on direct software interaction suggests a clear trajectory for the industry toward more integrated and autonomous AI assistants in professional environments.

Anthropic is aggressively commoditizing its frontier AI capabilities, pushing its most advanced features down-market into the Sonnet line to accelerate enterprise adoption and challenge the cost-performance ratio of competitors' flagship models.