n8n Labs Launches Nod, a Local-First Personal AI Agent for Mac and Windows

n8n Labs has introduced Nod, an open-source personal AI agent designed to run on a user's own computer rather than as a shared team automation platform. Released as a research preview on November 10, 2026, Nod is available for Mac and Windows, needs no n8n account, and operates in its own browser sandbox. The product signals a distinct direction from n8n's established workflow platform: a local-first agent for individuals who want to delegate routine digital work while retaining control of their data.

According to n8n Labs' official Nod announcement, users can give the agent a job in plain language, connect an AI provider or use a local open-source model, and let Nod work across websites and consumer applications. Its intended audience is solo operators and freelancers, although the approach may also interest business owners who need to test agent-led task automation without immediately building a multi-user deployment.

What Nod does and how it differs from n8n

Nod is a standalone personal agent, not a replacement for the core n8n workflow product. It runs in an on-device sandbox with its own browser and is designed to carry out work on behalf of an individual user. The published examples include cancelling subscriptions, watching prices, searching for or booking hotels, and retrieving data from dashboards.

The agent can interact with services such as Gmail, Calendar, Drive, Slack, and websites. That matters because many routine tasks span several browser tabs and applications but are too irregular or personal to justify building and maintaining a conventional workflow. Nod's research-preview model is to accept a job, execute it in the local environment, and ask for permission when an action falls outside the scope of the original request.

Key elements of the initial release include:

• Local execution in an on-device sandbox, with a separate browser for agent activity.
• Flexible model connections, including API keys, a ChatGPT plan, or open-source models run locally.
• Local retention of memory, logs, and saved credentials on the user's device or private server.
• Parallel jobs and scheduling for handling more than one task or recurring work.
• Cost controls that acknowledge model usage costs from the AI provider a user connects.
• Apache 2.0 open-source licensing, with code available on GitHub.

Nod is available for macOS 14 or later and as a Windows 11 preview. n8n Labs also says Nod can be hosted on a user's own server, while managed hosting and additional offline configurations are part of its roadmap.

The distinction is important. A repeatable business process can benefit from the structured logic and application connections associated with workflow automation. Nod instead focuses on a user's delegated, browser-based work, including tasks that may vary from request to request. The two approaches can be complementary, but n8n Labs presents Nod as a separate product and environment.

Privacy is central to Nod's design

Nod's local-first architecture is its most consequential product choice. The product says that its memory, activity logs, and saved credentials remain on the user's machine or private server. When an external model service is used, Nod sends only the data needed for the individual task to that service.

This does not eliminate the need for care. Users still choose which model provider to connect, what tasks to delegate, and what permissions to grant. A request that involves inboxes, calendars, cloud drives, or financial and travel sites can expose sensitive information depending on the task and provider configuration. Nod's permission checks are therefore a practical control: the agent should request approval for actions that were not explicitly authorized in the original instruction.

For a freelancer or owner-operator, this model can make experimentation more accessible. Rather than transferring all agent activity into a shared hosted system, they can evaluate a task on their own device, inspect what happened through local logs, and decide whether the result is reliable enough for continued use.

Where Nod could be useful, and where to be cautious

Nod is best understood as a research preview, not as an assurance that every browser task will be completed reliably. Its early value lies in testing work that is repetitive, time-consuming, and bounded enough for a user to review the outcome. A user might ask it to retrieve information from a dashboard, monitor a price, or prepare research across multiple sites.

The preview also makes model choice a practical consideration. API-connected models can incur provider usage costs, while local open-source models may offer a different setup and performance trade-off. The supplied research does not establish which model is best for any specific job, so testing should begin with low-risk tasks and clear approval boundaries.

Businesses should also avoid treating a personal agent as a substitute for a durable operational process before evaluating the task carefully. Work that needs consistent handling across people, systems, or customer records may require a more structured automation design. Nod's local agent approach is most compelling where the user wants personal control, browser access, and a lightweight way to explore what an agent can handle.

Need to turn successful personal-agent experiments into dependable operations? Scalevise helps businesses design practical n8n workflows that connect applications, reduce manual handoffs, and keep recurring processes maintainable as work grows. Our n8n workflow automation setup service can help identify which tasks belong in a personal agent and which need a structured, monitored workflow. Discuss your n8n automation project with Scalevise.

Frequently Asked Questions

What is Nod from n8n Labs?

Nod is a standalone, open-source personal AI agent from n8n Labs. It runs locally in an on-device sandbox, uses its own browser, and is released as a research preview for Mac and Windows.

Do I need an n8n account to use Nod?

No. n8n Labs says Nod does not require an n8n account to use.

Where does Nod store data and credentials?

Nod stores its memory, logs, and saved credentials on the user's machine or private server. When a connected model service is used, the product says it receives only the data needed for the task.

Which AI models can Nod use?

Users can connect their own AI provider through API keys or a ChatGPT plan, or run open-source models locally.

Is Nod available on Windows and Mac?

Yes. Nod is available for macOS 14 or later, and n8n Labs offers a Windows 11 preview.

Conclusion

Nod gives n8n Labs a distinct local-first personal-agent offering alongside its core workflow platform. Its open-source model, local storage approach, separate browser environment, and permission checks make it a notable research preview for individuals exploring delegated AI tasks. The practical next step is not broad automation, but careful testing of bounded work where users can review results, permissions, and model costs.