Hello everyone!
We’ve just launched the alpha version of our GPT-4o-Vision integration. As long as you select GPT-4o in the model choices in the builder, and make sure to check the box to allow end user upload of files, end users will now be able to upload images, just like on OpenAI.
We’re planning to add many more visual improvements in the coming weeks to the feature, but we encourage you all to give it a try!
That’s awesome. I just tested it with a pickaxe prompted to be an image expert. I loaded an image of some miso soup I made yesterday and asked it to create a similar image but as a painting. It gave me a beautiful image, much nicer than my original photo. It got everything right, from the peas to the spinach. I know it may be a silly use for it, but this image expert is part of my plan for using pickaxe, so in this case, it works perfectly. I can only upload one image, so I can’t show you the original, but the image, bowl, and everything else are there.
