I like to create image content descriptions (long text, captions, tags) using computer vision AI models and store the descriptions in Box metadata. Therefore i need to provide the image to an external service. The problem is, that the AI services available through e.g. Box Skills or most other 3rd party tools have a size limit of 20 MB per image. Most images i'm dealing with are much bigger. One option would be to convert these images prior to providing them to an external tool, but Box already has a smaller image which could be used for computer vision purposes – the file which is used in the preview (not the thumbnail). It would be great to get access to this preview file (by API, or by download from the preview screen).