How to link Image Recognizer to my Custom Vision project?

Questions about Thunkable
,
1

I have a Custom Vision project (which is a part of Cognitive Services by Microsoft) and I would like to link it to the Image Recognizer component of Thunkable. My project provides 2 keys: a Training Key and a Prediction Key. Which one do I use?

2

Did you can do this? I try do this, but something wrong with api