When I tried to use Gemini2.5 Pro API to develop a picture detection app, the uploaded pictures were always resized to an unknown size which is confusing me a lot. Even though I used prompt to configure its resize ratio, it did not work. Is there anyone who know the reason and could help me out.
You may either show us the code or ask developers online on websites like Stack Overflow.
gemini api doesnt have a way to configure to resize.
the images uploaded should be converted into respective embedding based on the embedding model used during training, so you cannot ask to resize the image.
also how could you tell if the image is resized ??