Multi-tool AI Workflow + Non-tool processes

Questions
, ,
1

Describe the problem/error/question

I have a Chat-based AI workflow, that uses a combination of Tools and (for file-based queries) separate HTTP node-based submissions to AI models. The issue I’m coming across is that when returning a response form a pre-defined AI node such as Gemini, the response returned to the Chat node is fine and correctly formatted. However when returning text from an HTTP node (which I’m having to do for submitting Audio files to Gemini as the Gemini node doesn’t support this), I get the JSON response printed in the chat. I’ve tried some Code node solutions from other threads (as in the segment of the workflow included here), but the response is always the same, e.g.:

{ “response”: "Okay, I’ll analyze the audio file you’ve provided!\nGeneral Overview:\nThis sounds like a quirky, almost cartoonish indie-pop or folk-pop song with a homemade, slightly lo-fi aesthetic.\nDetailed Analysis:\n• Genre(s): Indie Pop, Folk Pop, possibly with elements of novelty music.\n• BPM: Approximately 136 BPM […]
}

Is there any way I can have just the text returned to the Chat node without the JSON ‘wrapper’ whilst also keeping Tool-based elements in place (not included in the workflow below for the sake of keeping it simple)?

What is the error message (if any)?

No error, just a data / formatting issue.

Please share your workflow

Share the output returned by the last node

Information on your n8n setup

  • n8n version: 1.81.4
  • Database (default: SQLite): SQLite
  • n8n EXECUTIONS_PROCESS setting (default: own, main):
  • Running n8n via (Docker, npm, n8n cloud, desktop app): Docker
  • Operating system: Ubuntu 23
2

Just to clarify - I’ve realised I can ‘get around’ the issue by having an AI Agent node just repeat whatever has been fed into it, and the response is then returned as required. Is there a better way to do this though, it’s a bit of a waste to have an AI parrot the response back just to get the formatting returned correctly.

Many thanks for any help offered! :slightly_smiling_face: