Hi 
Wondering if anyone has found a way to take a PDF and extract the pages into images that can then be fed into OpenAI Vision?
Thanks
For document conversion, try these apps listed here https://www.make.com/en/integrations/category/file-conversions?addOnApps=1&nativeApps=1
e.g.:
2 Likes
Ah fantastic thanks! If the PDF is multiple pages will it output multiple PNGS?
You should give all of them a try and see which ones best suit your requirements.
2 Likes
Loving CloudConvert!! Very cool but how do you take a multipage PDF and export multiple JPGs? I cannot seem to see where to select that?
After converting, you need to “Get a Job”
(which can contain multiple tasks)
Then iterate the results files for the images
(where task operation is of type “export/url”)
And download the file before uploading it to Google Drive.
Using this I’ve managed to split an example PDF with two pages into two output images, and upload it to the same folder as the original file.
Fun fact:
Every 100 DPI in the original PDF is equivalent to 100% of the output dimensions of the image.
So if the PDF is in 300 DPI, the output image will be 3x the dimensions of 100% zoom of the PDF.
If you want to save on Make data transfer, try converting to JPG image output instead of PNG.
3 Likes
You are a champion!!! Thank you so much will try this out!!
Thanks @samliew ! Still got some errors:
I added everything as you did but it just stops at the iterator. Don’t quite understand the formula so hard to figure out why lol
You misspelt operation as operations.
2 Likes
Eek… and another word it seemed. Thanks so much it’s working now! How did you figure this out lol? Thanks again!
You just gotta play around and inspect the output bundles to get what you need. Sometimes the documentation doesn’t tell you enough.
2 Likes
yeah wow. ok great thanks. You should teach a course, I would sign up!
1 Like