Cannot parse JSON in HTTP Request module

Hi n8n community, It’s my first post here. Probably I am missing something obvious, but I am considering switching Make.com to n8n.

I am testing the AI agent from n8n, and I want it to do a website scrape from a service called jina.ai that typically returns a JSON.

However, in n8n, I cannot seem to get the output of the HTTP module to return a JSON, from which I can select only the section that is most relevant for the AI.

As you can see the JSON returns “response” and big chunk of monolith text, that is not separated into different objects.

Here are the request headers I am using:

At the same time, if I try building the same HTTP request with make.com I get the parsed JSON with the same settings.

I am sure I am doing something fairly obvious wrong, but what could it be? I couldn’t find a simple explanation neither through the AI Support Bot nor in the other threads in the forum. The solutions suggested were very convoluted.

Please share your workflow

Share the output returned by the last node

Information on your n8n setup

  • **n8n version: 1.76.1
    • **Database (default: SQLite): Default
  • **n8n EXECUTIONS_PROCESS setting (default: own, main): Default
  • **Running n8n via (Docker, npm, n8n cloud, desktop app): Cloud
  • Operating system: Mac

PS: I love my experience with n8n overall so far. I am glad I have found this tool! Looking forward to building more automation with it.

It looks like your topic is missing some important information. Could you provide the following if applicable.

  • n8n version:
  • Database (default: SQLite):
  • n8n EXECUTIONS_PROCESS setting (default: own, main):
  • Running n8n via (Docker, npm, n8n cloud, desktop app):
  • Operating system:

I’m still new to n8n so there may be a more efficient way but what I do is use the “Edit Fields” node to turn a string into a parsed Array whose fields can then be called later on.

I am also coming from Make and unfortunately, this feature isn’t natively built in as far as I can tell. The only node I’ve seen it in is the OpenAI message node where you can switch on “Simplify Output” and “Output Content as JSON”

1 Like

Hey!

THank you very much for commenting!

This makes sense. I expect that there are modules to handle this.

However, I was hoping that the module itself can handle this, as there is a section called “Optimise Response” and there is a field “Expected Response Type”, where you can select “JSON”

For some reason, when I select that, this doesn’t seem to be parsing the JSON correctly and I have no idea why.