Doctor Debug Join waitlist
n8n

A Chat Model sub-node must be connected and enabled

What this means

Your AI Agent has no 'brain' attached. An AI Agent node can't run on its own — it needs a language model (like an OpenAI or Anthropic chat model) plugged into it, and that model node has to be turned on.

Why you're seeing this

  • No Chat Model sub-node connected to the AI Agent

    Common

    The AI Agent was added (or its model was removed) without a chat-model sub-node wired into the 'Chat Model' connector underneath it. This is the primary cause.

  • A Chat Model is connected but disabled

    Sometimes

    The model sub-node exists but was deactivated (greyed out), so the agent still sees no usable model.

How to fix it

  1. 1

    With the AI Agent node open, click the '+ Chat Model' button at the bottom of the screen. With the node closed, click the 'Chat Model' + connector underneath it.

  2. 2

    Pick a chat model provider (e.g. OpenAI Chat Model, Anthropic Chat Model) from the list n8n shows.

  3. 3

    Open that chat-model sub-node, add or select valid credentials for the provider, and choose a model.

  4. 4

    Make sure the chat-model sub-node is enabled (not greyed out). If it's deactivated, right-click it and choose 'Activate'.

Why this happens (the technical detail)

n8n's AI Agent is a 'root node' in a cluster: it orchestrates but delegates the actual text generation to a connected Chat Model sub-node. Before running, n8n checks that at least one enabled node is connected to the agent's required Chat Model input; deactivated sub-nodes don't count. If none is found, it stops with this error before any API request is made.

Sources

Last verified 2026-09-17.