Anthropic has launched Claude Duet, a new AI agent designed to proactively assist professionals and small business owners. The agent is currently available as a research preview in the United States for Claude Pro and Max subscribers via a waitlist.
Claude Duet operates as a proactive agent, distinguishing itself from traditional chatbots by initiating tasks and providing timely information. It generates a daily briefing from connected applications, monitors them for urgent matters, and alerts users only when attention is required. The agent is built on Anthropic's Claude Opus 5.5 model and can handle a range of tasks, from scheduling meetings and drafting emails to building financial forecasts and creating web applications. It continues to work even when users are offline and prompts for approval before taking significant actions, such as sending messages or making payments. This proactive approach aims to help users prioritize their work and dedicate more time to complex projects.
The agent is accessible through a standalone app on iOS and Android devices, as well as via the web on claude.ai. Anthropic emphasizes that Claude Duet carries over memory, skills, and existing app connections from a user's Claude account, allowing it to begin with established context. During the research preview, Duet's usage counts towards the existing Claude plan limits, with more frequent background checks consuming allowances faster. Anthropic has stated that safety and discretion are core to Duet's design, drawing on their framework for trustworthy agents, though further details on specific safety protections are expected. The company has not yet detailed plans for Duet's availability on Team or Enterprise subscriptions, nor has it announced an API for broader integration. This launch positions Claude Duet as a competitor to similar agent offerings from Meta and OpenAI, such as Muse and Dots, respectively.