←News detail

WallstreetcnUpdated Oct 25 00:15

Anthropic launches Claude Duet, an around-the-clock AI agent powered by Opus 5.5

AI summary

Anthropic has launched a new app called Claude Duet, positioned as an AI agent that can run continuously in the background, proactively identify needs, and execute tasks, built on Claude Opus 5.5. The product comes with an independent virtual machine and browser environment and can call on more than 6,000 connectors and plugins in the Claude Marketplace, covering services such as Gmail, Slack, QuickBooks, and Stripe; key steps such as sending messages, making payments, and signing documents require user approval. It is currently in research preview and available only to Claude Pro and Max subscribers in the United States through a waitlist, and Anthropic also acknowledges that it still has limitations in judging how urgent information is and handling personal affairs.

Why read this

AI score 73

The original text explains background execution, more than 6,000 connectors, and the approval mechanism, so readers can judge the capability boundaries and limitations of proactive agents.

Topics

Tags

Body · AI translation

Anthropic is extending AI competition from model capability to task execution.

On Wednesday local time, the company launched a new app, Claude Duet, positioning it as an AI agent that can run continuously in the background, proactively identify user needs and execute work tasks, further expanding the product form of Claude.

Unlike traditional conversational assistants that wait for users to issue instructions, Claude Duet can aggregate important information from connected apps, proactively send notifications when user attention is needed, and handle a variety of tasks ranging from meeting scheduling to financial forecasting and web app development.

However, the product is currently still in research preview and is only available through a waitlist to Claude Pro and Max subscribers in the United States. The limited availability means that Claude Duet has not yet entered a full competition phase targeting a broader user base.

Anthropic said Claude Duet is built on its flagship model Claude Opus 5.5 and can continue executing tasks after the user leaves the device, but it still has limitations in judging the urgency of information and handling some personal matters. As Meta and OpenAI also advance proactive AI agents, Anthropic is trying to turn model capabilities into user services.

From responding to instructions to proactive execution

According to an official announcement released by Anthropic, Claude Duet is designed for professionals and small business owners with heavy workloads.

The app summarizes important items from connected apps every morning and continuously checks relevant information throughout the day, proactively sending reminders when it finds matters that require user action.

Users can also directly assign tasks. In addition to routine administrative work such as scheduling meetings, Claude Duet can also use multiple tools to build dashboards, create financial or sales forecasts, and develop prototypes and web apps.

The core change is that users do not have to stay in the conversation interface and issue instructions step by step; the agent can advance tasks in the background and notify the user when they are completed or need review.

Anthropic said Claude Duet inherits the memory, skills and connected apps in the user's Claude account, reducing the need for repeated configuration. The product offers standalone iOS and Android apps and can also be used via the web.

Built on Opus 5.5, connecting more than 6,000 tools

Claude Duet is based on Claude Opus 5.5 and is equipped with an independent virtual machine and browser environment, enabling it to continue working after the user's computer or phone is turned off.

According to Anthropic's introduction, the product can call more than 6,000 connectors and plugins in the Claude Marketplace, covering work tools such as Gmail and Slack as well as business services such as QuickBooks and Stripe.

These connectors enable the agent to combine information retrieval with follow-up actions. For example, it can check unprocessed emails and unpaid invoices, draft reminder messages, prepare payroll processing content, and also assist in creating marketing materials, managing online stores or preparing contract documents.

Key steps such as payments and sending messages remain subject to an approval mechanism. Anthropic said that before sending messages, making payments, signing documents or deleting information, Claude Duet will show the user the proposed action and wait for approval.

The product also has dedicated security mechanisms to prevent improper disclosure of sensitive information; some high-risk operations cannot be executed even with user approval.

Research preview stage, capabilities and scope of use still limited

Claude Duet is not currently a fully open commercial product. This time, Anthropic first opened waitlist applications to Pro and Max subscribers in the United States. During the research preview, the agent shares subscription usage quotas with other Claude features, and background checks and complex tasks also consume quota.

Anthropic also acknowledges that at this stage the product may still misread conversation content or misjudge the urgency of matters. Because the product is mainly developed around work scenarios, its ability to handle personal matters such as booking tickets is relatively weak, and users currently cannot directly view its browser operations.

Therefore, the actual value of Claude Duet still depends on whether it can form a stable workflow among continuously obtaining information, proactively judging task priorities and executing operations. For enterprises and professional users, background automation is expected to reduce information screening and repetitive operations, but misjudgments and the handling of sensitive information remain issues to watch during use.

As Anthropic launches Claude Duet, OpenAI and Meta are also advancing AI agents that can execute tasks on behalf of users. OpenAI previously launched Dots, and Meta launched Muse; both products attempt to move AI beyond answering questions toward completing tasks across apps.

The Verge sees Claude Duet as a competitor to Meta Muse and OpenAI Dots, and notes that Duet integrates memory, skills and app connectivity in the Claude ecosystem, executing work tasks through a virtual machine and browser.

Business Insider noted that Claude Duet is still in the waitlist opening stage and cannot yet immediately compete directly with other agent products in the broader market.

For Anthropic, Claude Duet is not only an expansion of the Claude app form, but also means the company is exploring how to keep models continuously involved in users' daily work, rather than only providing one-off Q&A and content generation services. Its subsequent development will depend on the reliability of task execution, users' acceptance of proactive services, and whether the product can gradually expand its scope of use from a limited research preview. Risk disclosure and disclaimer

Markets carry risks, and investment requires caution. This article does not constitute personal investment advice and does not take into account the specific investment objectives, financial situation or needs of individual users. Users should consider whether any opinions, views or conclusions in this article are suitable for their particular circumstances. Investment based on this is at the user's own risk.

Source

Wallstreetcn