Hi Retool forum,
I'm testing a new prototype using Retool AI, but I'm running into issues using Vision. When I use fileDropzone to upload a test image to test generating text of an image, the output returns an error "400 invalid image". I've tried different files types (e.g. png, jpg), different file sizes (e.g. 500kb to 5mb), and encoding then decoding the image, but I have not been able to get around this bug. Anyone experienced this before have an answer?