n8n

Passing everything from one LLM step to the next can quickly fill your context window — and hurt performance. Our latest article covers 6 practical ways to manage context across multi-step LLM workflows, from keeping workflow state to setting token budgets. See how to put them into practice with n8n 👇

How do you scale automation without bringing chaos to your company? In The Loop will present three separate sessions from Bitovi, EdDev, and Bayer with Avanai, on the governance models that make that possible. You’ll find out: 🎙️How to prioritize use cases and train teams through a Center of Excellence model, with Mick McGrath 🎙️ How to scale n8n inside regulated enterprises with shared governance and claim 3,000 hours back, with Ric Neves and Marcelo França 🎙️How to share memory, governance, and logic across agents, with Christian Voigt Ready to let more people build without losing control of security, standards, or compliance? Join us in Berlin on 14 October: link in the comments 👇 #InTheLoop2026

This new featured template checks your product listings on a schedule, picks up new reviews, and reports the sentiment along with a summary of what customers said. You get a Telegram message when something lands, and a full log in Google Sheets for tracking trends over time. It works just as well pointed at a competitor's listing. Check it out in our Template Library: https://bit.ly/4hkQZ2y

Connecting a chatbot to your internal APIs is the easy part. Making sure it only accesses and changes what it should is harder. Our new guide breaks down chatbot API integration in 6 practical steps — from defining tools and verifying user identity to permission checks, human approval, safe API responses, and audit logs. See how to build secure tool calling between your AI model and internal APIs with n8n 👇

The closest thing to a deterministic LLM call just landed in n8n. And it's free via n8n Gateway credits until Oct 10. TypeSafe AI's Jev is the model everyone's been talking about this month. It doesn't write text. It doesn’t hallucinate. You give it an input, a question, and your desired outcomes and it returns the probability for each one. Low latency and low cost mean you can use the TypeSafe AI node anywhere a workflow needs a judgment call: → Does this deal need an exec review? → Is this ticket critical, high or low priority? → Complaint, question or spam? 🎁 On n8n Cloud (Starter, Pro, and new trials), Jev runs on Gateway credits and it's free until October 10. If you’re self-hosted or on the Community Edition, you can BYOK once the node is installed. Search "Jev" in the nodes panel (refresh your instance if you don't see it yet) to try it in your own workflows. 🎥 Want to see it in action? Check out Liam’s video below. https://go.n8n.io/FsedIr

n8n now has a new kind of agent, and it's super easy to set up 🤖 In this video, Liam builds one from scratch: a personal assistant you can text on Telegram. No code, no workflows. Just instructions, a few tools and a daily schedule. By the end, it turns a photo of a sticky note into a Todoist task, logs a receipt as an expense, and goes through yesterday's email every morning at 9 to pull out the to-dos. Watch the full walkthrough 👉 https://lnkd.in/ehZh3S6v

When we started planning In The Loop, it was clear that we should hear directly from the teams who use n8n in their businesses. Now Mercedes-Benz, Bosch, Webflow, and Trade Republic are joining us in Berlin on October 14. They will share: - How Mercedes-Benz AG runs n8n in production - How Bosch is engineering AI agents - How Webflow powers agentic workflows - How Trade Republic has built and extended n8n with custom nodes I am especially excited for these sessions because they are a chance to hear what is actually happening beyond the demos: how teams are building, scaling, and adapting n8n to fit the way they work. I would love to see you in Berlin on October 14 for our first In The Loop. https://n8n.io/intheloop/