Anthropic has opened a waitlist for Claude Duet, a research preview that lets its Claude assistant keep working while users are away, according to Tech Crunch.
The preview is available to Pro and Max subscribers in the US through iOS, Android and claude.ai, included in existing subscriptions at no extra cost, though Duet's activity counts against the same usage limits as other Claude features.
Duet runs on Claude Opus 5.5, the top tier of Anthropic's model lineup built for complex, long-horizon agentic work leads in agentic coding and knowledge work, and operates inside its own virtual machine complete with a browser, so users do not need to keep a device running for it to operate.
Connected apps
The agent checks connected apps every few hours, sends a morning summary and flags items needing a decision, carrying over memory, skills and integrations from a user's existing account; Anthropic says it supports more than 6,000 integrations including Gmail, Slack, QuickBooks, Stripe, HubSpot and Shopify.
Sensitive actions still require sign-off: Duet shows its plan and waits for approval before sending messages, making payments, signing documents or deleting files, and it blocks securities trading and account-to-account transfers outright, though purchases can go through Stripe's Link wallet or Shop Pay via a virtual card.
Early reviews mixed
Early reviewer Dan Shipper of Every found Duet's email drafts weaker than a rival agent's and noted its voice capability is "There's no real voice mode yet, just transcription," he wrote.
The launch lands amid what one research brief frames as a shift from chatbots that wait for prompts to proactive agents that monitor information and interrupt only when needed, with OpenAI's Dots and Meta's Muse pursuing the same ground.
Anthropic, which closed a compute deal valued near $45 billion, said it would share more detail on Duet's safety design in the coming days.
Read more about AI Models & Research
