Adding a file to the request does not work

[CeleritasCelery]

openai-image-variation-prompt does not currently work. it is just passing the filename as a parameter instead of attaching the image itself. I tried to fix this with the code below:

(openai-request "https://api.openai.com/v1/images/variations"
    :type "POST"
    :headers `(("Content-Type"  . "application/json")
               ("Authorization" . ,(concat "Bearer " openai-key)))
    :data (json-encode
           `(("n"               . ,openai-image-n)
             ("size"            . ,openai-image-size)
             ("response_format" . ,openai-image-response-format)
             ("user"            . ,openai-user)))
    :files `(("image" . ,image))
    :parser 'json-read
    :success (cl-function
              (lambda (&key data &allow-other-keys)
                (funcall callback data))))

However this returns an error from openai

Invalid Content-Type header (application/json), expected multipart/form-data. (HINT: If you're using curl, you can pass -H 'Content-Type: multipart/form-data')

So it looks like it doesn't like mixing form data with json encoded data (and you can't json encode images). I tried to switch to using only form encoding using this:

(openai-request "https://api.openai.com/v1/images/variations"
  :type "POST"
  :headers `(("Authorization" . ,(concat "Bearer " openai-key)))
  :data `(("n"               . "1")
          ("size"            . ,openai-image-size)
          ("response_format" . ,openai-image-response-format)
          ("user"            . ,openai-user))
  :files `(("image" . ,image))
  :parser 'json-read
  :success (cl-function
            (lambda (&key data &allow-other-keys)
              (funcall callback data))))

But that got a different error from openai

400 - Additional properties are not allowed ('--form image' was unexpected)

I am not very familiar with HTTP so I am stuck on how to get this working. Any help would be appreciated.

[jcs090218]

Yeah, I am going to rewrite the library using the better approach, see #5. Sorry for this isn't being 100% ready. 😓 I feel sorry since I created package for my own use, but I will soon make improvements next week! 😅

[CeleritasCelery]

I didn't realize this was still "beta". Amazing work! I was really impressed and I thought I would have to implement something from scratch.

[jcs090218]

Thanks! It's not great yet, but I will continue to make improvements! I'm busy for the following two weeks, I will try to work on it the next weekend! :D

[CeleritasCelery]

Are you open to PR's or are you planning on doing a complete rewrite?

[jcs090218]

I've made quick adjustments (#7) so now all request parameters are all exposed! I was hoping to expose response with class or structure, but unfortunately, I don't have the knowledge to do that now. 😅 I wasn't comfortable asking people to make PR before a rewrite, but now that I've made the adjustment, feel free to open PRs! :D

[jcs090218]

Hi, I've recently tried the request but I'm encountering the same issue! After some time of google searching, it seems like there are still some issues with the image requests. I will mark this as won't fix for now since there is no way we can fix it on our side!

[CeleritasCelery]

What do you mean by "there are still some issues with the image requests"? Are you referring to a limitation in the request.el library? If so I don't see an issue for this in their tracker. However the image API works fine over curl, so it is not a limitation on the OpenAI side. For example this will work:

curl  https://api.openai.com/v1/images/variations \
     -H 'Authorization: Bearer <your key>' \
     -F "image=@</path/to/image.png>" \
     -F n=1 \
     -F size="256x256"