Cannot retrieve AI response from API — POST works, but GET returns 404

[test1]

Hello everyone,

I’m trying to use n8n to interact with the 1min.ai API to analyze an image and generate Instagram hashtags. Here’s my current setup:

1. I send a POST request to https://api.1min.ai/api/conversations with:

• An image (passed in the body as base64: data:image/jpeg;base64,...)
• A prompt asking for exactly 5 Instagram hashtags
• Required headers: API-KEY and Content-Type: application/json

The request works: I get a 200 response and a conversation UUID.

But the response does not contain the AI-generated text (e.g. hashtags). It only returns the conversation metadata.

2. I then try to retrieve the AI response using:

GET https://api.1min.ai/api/conversations/:uuid

or

GET https://api.1min.ai/api/conversations/:uuid/messages

Result: 404 - Cannot GET /api/conversations/...
Even though the POST was successful.

---

What I’ve tried:

• Double-checked my API-KEY — it’s correct.
• Verified the image is properly passed in base64 ({{$input[0].binary.data.data}}).
• Tested the same POST in Postman — same result: no response in the body.
• Confirmed the image is received (no error from API).
• Waited a few seconds and retried GET — still 404.

---

My questions:

1. Is it possible the API is asynchronous, and I need to poll for the response?
2. Could the AI response be returned directly in the initial POST — but only if the image is sent differently (e.g. multipart/form-data)?
3. Has anyone here successfully used 1min.ai with n8n? Or faced a similar issue with APIs that create resources but don’t expose them via GET?

Any help or debugging tips would be greatly appreciated!

Thanks in advance!

[Olayiwola_Maryam]

Hi test1

You’re likely dealing with two main issues here:

1. Asynchronous API Behavior

It seems the 1min.ai API creates a conversation but processes the AI response in the background. That means you get a UUID immediately on POST (success),
But the AI-generated hashtags are not ready yet, and
If you try to fetch them immediately, the GET /conversations/:uuid/messages returns a 404.

Solution you can Implement
Add a delay (wait 3–5 seconds) before making the GET request to check for the response.

Data Format: base64 vs. multipart

Some APIs expect multipart/form-data instead of base64 in JSON for file uploads.

Solution to Try:
If the API documentation allows multipart/form-data, try sending the image As a binary file (using n8n’s HTTP Request node in multipart/form-data mode),
Not just a base64 string

This might ensure the image is processed correctly and triggers the AI to generate hashtags.

Ensure Correct UUID Usage

Double-check that you’re:

Not accidentally including /api/conversations/:uuid with a placeholder instead of the actual UUID.
Not missing any required headers in the GET request (some APIs need the API-KEY on every call).

If you would like, I can jump on a quick call and help you set this up step-by-step in n8n including delay logic, correct headers, and multipart handling.

Let me know. I am happy to help troubleshoot it live.