This is a cool app that will describe anything you don’t have the words to describe yourself. For instance, my cat. I took a picture of my cat sitting on a chair and it was able to describe everything about the image nicely. What is very useful is being able to use the descriptors for use in ChatGPT or another AI model for creation purposes. One request I have that prevents me from giving 5-stars is that I can’t find a way to copy and paste the output from the app. Right now I take screenshots of each paragraph, open Photos, and copy/paste each text into a document. Please add the ability to copy and paste or perhaps have a “history” where it saves the conversations for later use!

I like the huggingface platform very much, so I gave it five stars, but why does it keep flashing back when I use it?

It’s a decent app to run visual models offline, but it falls short in several aspects. Every time I reopen the app, it gets stuck initializing the model and then proceeds to download the models again. Additionally, the offline mode is not well-supported. When there’s no internet connection, it displays an error message indicating “no internet connection.”

After trying about 10 to 15 images across various scenes, I could not use this for any need that required accuracy. The idea is great and it does describe a lot of things, but the accuracy of the items it describes are almost all completely wrong. For example, it called a couch, a floor, and a wooden serving tray, a gaming console because there was a controller sitting on it. Hopefully they have more time to develop a more accurate product.