Hi everyone,
I finally managed to download product images from our Shopify store into Make.com using the HTTP → Get a file module.
The goal is to upload these images to ChatGPT so it better understands our products and can generate realistic lifestyle images based on them.
The HTTP → Get a file step works fine — I see the binary files in Make.
However, when I tag those files in the ChatGPT (OpenAI) module prompt, the output log shows "buffer" instead of the image content, and the generated image doesn’t visually match our actual products.
I’ve also tried using the “Analyze images (Vision)” module, but the results haven’t been very accurate either.
My question:
How can I correctly pass image files from Make.com to ChatGPT so they are actually used as reference images in generation?
