this tutorial is demonstrating an alpha version of n8n AI nodes. We removed this node shortly after and merged the functionality with “Basic LLM Chain” node. So to implement structured output parsing, you’d need to enable “Require Specific Output Format” option and then connect “Structured Output Parser”.
In the sub-node you can either chose if you’d wish to provide an actual JSON schema or an example JSON object which would be converted to schema in the background. Here’s a simple example workflow: