Nod by n8n

Key facts: Nod launched on Product Hunt on November 10, 2026, from n8n labs. It is an open-source personal agent with its own computer: it chases invoices, preps your calls and cancels trials, on real websites, on your machine, and it asks before it acts for you. In n8n's own tests, 678 runs finished the job 98% of the time, remembered what it was told 93.2% of the time, and fell for 0 of 216 published attack emails trying to plant a false memory. Errand types come from 900,000+ anonymised production workflows. It runs on macOS 14 and later, with a Windows 11 preview, under Apache 2.0.

Plans: Free and open source (Apache 2.0). Use ChatGPT Plus or Pro, or any AI key: you pay your model provider, not n8n, for the AI itself.

Best for: Anyone with more errands than time who wants an agent that works on real websites but never acts without permission. Who should skip it: anyone who wants a fully autonomous agent that never asks, or who has no AI plan or key to run it on.

Pros

• Free and open source under Apache 2.0
• Asks before it acts: zero actions without permission in testing
• Tested against prompt-injection attacks: 0 of 216 got through
• Runs on your machine, on real websites, not a sandbox demo

Watch out for

• You still pay for your own AI plan or key
• Windows is a preview: macOS is the finished build
• An agent on real websites needs real trust to set up

Best for

Anyone with more errands than time who wants an agent that works on real websites but never acts without permission. Who should skip it: anyone who wants a fully autonomous agent that never asks, or someone with no AI plan or key to run it on.

Pricing

Free and open source (Apache 2.0). Runs on ChatGPT Plus/Pro or any AI key, billed by your model provider.

FAQs

What is Nod? An open-source personal agent from n8n labs, with its own computer: it chases invoices, preps your calls and cancels trials on real websites, and asks before it acts.

Is Nod free? Yes. It is free and open source under Apache 2.0. You pay your model provider for the AI itself, through ChatGPT Plus/Pro or any AI key.

Does it act without asking me? No. It asks before it speaks for you: in testing, 678 runs took zero actions without permission.

Can a scam email trick it? In n8n's tests, 0 of 216 published attack emails managed to plant a false memory.