How to use n8n Chat Hub to run workflows in chat

TL;DR

The Chat Hub lets you run workflows and agents directly in chat, replacing older custom GPTs and aggregators. You can pick multiple models, turn on tools like web search and voice control, and enable custom agents that invoke workflows with a simple toggle. This centralizes chat and automation in one place for easier back and forth with LLMs.

Transcript

If you're still using custom GPTs like his 2023 or something like typing mind or PO, you're going to want to check this out. Niden just released something that replaces all of them and adds on one more important feature, the ability to execute your workflows and your agents all in one place. So, not only can you chat with things like Claude, Gemini... Read More

Key Insights

• n8n Chat Hub consolidates multiple LLMs and workflows into one chat interface.
• Tools like web search and image uploads are available in the chat hub and can be toggled on.
• Custom agents are the central feature, allowing workflows to be invoked from chat with a simple selection.
• Workflows can be invoked via chat using a specific agent and a text input such as a URL or transcript.
• The example web scraper agent demonstrates invoking a Firecrawl based workflow from within chat.
• To enable a workflow to appear in the hub, you must publish it and enable the make available in any chat toggle.
• Self-hosted users may need to add a brand new chat trigger to expose the workflow in the hub.
• If a workflow does not appear, rebuilding the workflow node in a new workflow is suggested as a troubleshooting step.

Questions & Answers

Q: What is the main benefit of the n8n Chat Hub compared to older custom GPTs

The Chat Hub centralizes talking to different LLMs and invoking automations in the same space, so you can chat with models like OpenAI, Anthropic, and Google while running your own workflows. This reduces the need for separate tooling, APIs, and schema management and provides a single place to trigger actions and access diverse models.

Q: How do you access the hub in the n8n app

To access the hub, you update to the latest n8n version and click the chat icon at the bottom left. This opens the hub interface where you can pick from different language models, including OpenAI and other providers, and access built in tools and custom agents.

Q: What models are supported in the Chat Hub

The hub supports OpenAI models as well as other providers like Anthropic, Google, and Open Router based options. You can switch between these models within the hub to leverage different capabilities and responses for your conversations and workflows.

Q: What are the built in tools available in the hub

Built in tools include web search, voice control, and image uploads. These tools can be toggled on to enrich conversations and enable the LLM to fetch information from the web, speak responses, or analyze uploaded media as part of the chat workflow.

Q: What is a custom agent in the Chat Hub

A custom agent is a workflow that you publish and expose in the chat hub so it can be invoked from within a chat. You name and describe the agent, publish it, and then you can trigger the workflow using natural language commands or chat input to perform tasks like scraping or data processing.

Q: How does the web scraper example work in the hub

The example shows selecting a web scraper agent, providing a website URL, and letting the agent invoke a Firecrawl based workflow behind the scenes. The hub then returns the scraped data, demonstrating how a chat based trigger can drive an automated page crawl and data extraction.

Q: What should you do if a workflow doesn’t appear in the hub

If a workflow does not appear, rebuild the workflow node by node into a brand new workflow and test again. If you still have issues, add a chat trigger to the workflow and verify that it can be invoked by natural language input, ensuring it is published and available across chats.

Q: Why is this Chat Hub considered a replacement for older GPTs and aggregators

The hub replaces older solutions by centralizing model access, tools, and the ability to execute your own automations in one place. It eliminates the need for separate OpenAPI schemas and webhooks for each workflow, enabling chat driven actions and back and forth with multiple LLMs without switching platforms.

Summary & Key Takeaways

• The video explains the new n8n Chat Hub and how it consolidates model access, tools, and workflow execution in a single chat interface, replacing older solutions like TypingMind and custom GPTs.

• It covers how to access the hub, select models, enable and use tools, and attach custom agents that can trigger Python-like or web scraping workflows, all through natural chat interactions.

• It also highlights self-hosted caveats, the necessary toggles to enable workflows in the hub, and troubleshooting steps for workflows not appearing in the chat hub.