I’m excited to announce that we’re releasing a brand new feature called Chat Hub, available on n8n 2.1.0 beta right now.
With so many foundational models in the AI market, one thing remains constant: they all have a very familiar chat interface. Whether you’re asking about vacation recommendations or coding a new user interface, using natural language to communicate with AI has become one of the most popular methods of conveying questions and ideas. Products like LibreChat and Langdock are examples of how AI chat is helping workers interact with business tools and data.
Chat Hub takes this concept to the next level, giving you a single chat interface that allows you to directly interact with n8n workflows, create your own agents, and access multiple AI models. This opens a huge amount of possibilities for how people can use Chat Hub.
Chat Hub introduces a powerful new way to give users access to your n8n automations without exposing the underlying workflows. You can assign users a “Chat only” role, giving them a ChatGPT-like interface to interact with your workflow agents—nothing more. These users have no access to see or edit workflows.
Within Chat Hub, you create workflow agents that define exactly how users can interact with the workflows you choose. And because the chat interface is built directly into n8n, you can offer this experience without relying on Slack or any other third-party tools.
Teams and businesses benefit from have a single point that all their users can access AI models and agents, controlling usage and costs under centralized credentials as well as ensuring data, access security and AI model usage are kept in check.
Be aware Chat Hub is in beta and will be in active development. Please follow our release notes page for latest updates: https://docs.n8n.io/release-notes
There is an issue with local n8n chat hub connecting to Local Ollama LLm, the api showing connected successfully and working on workflow but not working on chat hub , getting error with all different models from ollama
This is a really solid update — Chat Hub looks like a big step forward for teams that want to expose AI-powered workflows without exposing the complexity behind them.
I’ve been working extensively with n8n, AI agents, and chat-based workflow interfaces, including designing “chat-only” user experiences, multi-model setups, and secure agent access for teams. I can help with:
Designing and structuring workflow agents for Chat Hub
Best practices for permissions, security, and cost control
Connecting Chat Hub to real business use cases (support, ops, internal tools)
Optimizing AI model usage inside n8n
If anyone wants help implementing or exploring Chat Hub in a real-world setup, feel free to DM me on Discord: peterpipedrivepro — happy to help or brainstorm use cases.
In version 2.1.4, when using openai credentials in the conversation, the Base URL configuration in the credentials was not used, resulting in MODEL_NOT-FOUND; The same credentials can be used normally in the AI agent node
Title generation failed: Error: 404 status code (no body)
Troubleshooting URL: MODEL_NOT_FOUND - Docs by LangChain
Hi team members. I’m new and probably the oldest here. Im from GEN Boomer. And i can see that im amongst the new GEN. But, I hope to learn more from you, guys. Looking forward to work hard
Pretty Cool, tried it out myself, works like a charm! Though I do have an issue, I have a workflow activated that has a chat trigger and would appear under Workflow Agents but mine says “No workflow agents available.” don’t know if this is still in the works?
Ok Forget this, I forgot to turn on the Make available in n8n chat option.
