Message GPT Model keeps getting error

Questions
1

Hi everyone,

I’ve been watching YouTube and recently started learning automation. I’m currently trying to follow his email automation/tutorial videos step by step.

I copied the workflow exactly as the youtuber showed in the video, but I keep getting errors that he doesn’t seem to get.

These are my errors:

image
image629×637 101 KB

image
image1080×549 89.8 KB

image
image1080×530 59.1 KB

I’ve already tried:
• Changing the GPT prompt
• Switching to different GPT models
…but the error still happens.

This is the youtuber’s method:

image
image529×437 62.6 KB

I’m still pretty new to this, so I’m wondering if I missed something obvious or if there’s an update/change since the video was posted.

Has anyone run into the same issue or knows how to fix it?

2

Hey @fer_not Welcome!
Can you share the workflow with us? Also for now try using the AI agent and there give your message 1 as a system prompt and your AI prompt as the message 2, and connect your OpenAI model there it should work that way.

3

image
image836×385 19.9 KB

Hi @Anshul_Namdev , thank you for the reply! The original workflow above.

I tried to do it as your recommendation, and it comes out with an error

image
image684×318 25.6 KB

image
image1565×904 147 KB

image
image305×173 7.69 KB

4

@fer_not can you do this?

image
image996×479 15.8 KB

6

welcome to the n8n community @fer_not
I believe the error is happening because the tutorial uses an older OpenAI node setup, and in the current n8n version you need to enable structured JSON output (or disable “Require Specific Output Format”) and make sure the prompt returns valid JSON instead of plain text.

7

Use a structured output parser instead.