Anthropic Launches Claude Sonnet 4.6 with New Skills

Anthropic has officially released its newest AI model, Claude Sonnet 4.6. This update comes just a few weeks after the launch of their premium Opus 4.6 model. For most people using the free version of Claude, this is the version they will interact with every day.

The headline feature of this update is the 1 million token context window. To put that in perspective, a token is roughly equivalent to a word or part of a word. Having a 1 million token window means the AI can “read” and remember thousands of pages of text in one go.

You could upload an entire series of technical manuals or a decade of company emails and ask the AI to find a specific detail. It is currently in a beta phase, but it represents a huge leap in how much information a computer can process at one time.

One of the most impressive parts of this launch is the focus on coding. Many developers rely on Claude because it tends to follow instructions more accurately than other models. Anthropic claims that “Claude Sonnet 4.6 is our most capable Sonnet model yet.” They specifically highlighted that it is a full upgrade for skills like coding, computer use, and agent planning.

Another big focus is “computer use.” This is a feature where the AI can look at a screen, move a mouse, and click buttons. While it sounds like science fiction, it is a practical tool for automating boring office work. Anthropic mentioned that “Performance that would have previously required reaching for an Opus-class model, including on real-world, economically valuable office tasks, is now available with Sonnet 4.6.”

For iPhone app users, the update should be live right now. The price for developers using the API remains the same as the previous version. It costs $3 per million input tokens and $15 per million output tokens. For the average person using the website, you just get a smarter, faster assistant for free or through your existing Pro plan.