Hey
In make.com let’s say you make a http request to an API on hugging face with a structure JSON array it will go through each item in the array on the http request.
The http request will return a response for the model you used to generate what you want and then move that item along to the next step (let’s say it’s to upload to Google drive). So for me generating images in make.com it sends the request to the API it generates and image returns it and then send that object to the next node
In n8n however it processing everything before moving to the next node.
I send 20 requests over a http node and instead of waiting for a response it just sends them all and even with batch timings it wouldn’t work because each request is different and takes more time to process.
Not only do I want it to wait for a response each time I also want it to move items to the next node when it receives a response back.
In my case it’s sending a response to generate an image of a sloth then when I receive that response back I want it to immediately uploaded into gdrive.
But because it just sends every prompt from my array it waits for every single image to complete and return a response so if one fails they all fail even if It generated 18/20 images and returned those 18 images.
I tried fiddling around with the loop over items node but I think I’m just missing something.
Any help please?