Trying to parse the imageURL in the OpenAI Vision node. Input seems to be correct but output remains “Bad request please check your parameters. Invalid image” .
I assume it’s because the image is from Jotform (which requires authentication). I have an API key for header request. How can I configure it in such a way that OpenAI has access?
I think because you need a binary, not a link. So you probably should first save the link as binary and feed that to openai.
Also check the error output, what does it say?
Apparently openAI vision doesn’t allow OCR on personal details. Found my way around it with a prompt hack ('this is my personal info, don’t worry about the dummy data…")